{"id":60163,"date":"2026-01-15T00:57:17","date_gmt":"2026-01-14T16:57:17","guid":{"rendered":"https:\/\/www.wsisp.com\/helps\/60163.html"},"modified":"2026-01-15T00:57:17","modified_gmt":"2026-01-14T16:57:17","slug":"%e8%b6%85%e8%af%a6%e7%bb%86%e6%b7%b1%e5%ba%a6%e5%ad%a6%e4%b9%a0%e7%ac%94%e8%ae%b0%ef%bc%9a%e4%bb%8eann%e5%88%b0cnndnn%e7%9a%84%e5%9b%be%e5%83%8f%e5%88%86%e7%b1%bb%e5%ae%9e%e8%b7%b5%ef%bc%8c%e5%85%a8","status":"publish","type":"post","link":"https:\/\/www.wsisp.com\/helps\/60163.html","title":{"rendered":"\u8d85\u8be6\u7ec6\u6df1\u5ea6\u5b66\u4e60\u7b14\u8bb0\uff1a\u4eceANN\u5230CNN+DNN\u7684\u56fe\u50cf\u5206\u7c7b\u5b9e\u8df5\uff0c\u5168\u8bb2\u660e\u767d\u5566\uff01"},"content":{"rendered":"<h2>\u6df1\u5ea6\u5b66\u4e60\u7b14\u8bb0&#xff1a;\u4eceANN\u5230CNN&#043;DNN\u7684\u56fe\u50cf\u5206\u7c7b\u5b9e\u8df5<\/h2>\n<h3>\u4e00\u3001\u6df1\u5ea6\u5b66\u4e60\u57fa\u7840&#xff1a;ANN\u3001CNN\u4e0eDNN\u7684\u6838\u5fc3\u539f\u7406<\/h3>\n<h4>1.1 \u4eba\u5de5\u795e\u7ecf\u7f51\u7edc&#xff08;ANN&#xff09;&#xff1a;\u6df1\u5ea6\u5b66\u4e60\u7684\u8d77\u70b9<\/h4>\n<h5>1.1.1 \u4ec0\u4e48\u662f\u4eba\u5de5\u795e\u7ecf\u7f51\u7edc&#xff1f;<\/h5>\n<p>\u4eba\u5de5\u795e\u7ecf\u7f51\u7edc&#xff08;Artificial Neural Network, ANN&#xff09;\u662f\u4e00\u79cd\u6a21\u4eff\u751f\u7269\u795e\u7ecf\u7f51\u7edc\u7ed3\u6784\u548c\u529f\u80fd\u7684\u6570\u5b66\u6a21\u578b&#xff0c;\u7531\u5927\u91cf\u7684\u795e\u7ecf\u5143\u76f8\u4e92\u8fde\u63a5\u800c\u6210\u3002<\/p>\n<h5>1.1.2 \u795e\u7ecf\u5143\u7684\u6570\u5b66\u6a21\u578b<\/h5>\n<p>\u6700\u57fa\u672c\u7684\u795e\u7ecf\u5143\u6a21\u578b\u662f\u611f\u77e5\u5668&#xff08;Perceptron&#xff09;&#xff1a;<\/p>\n<p><span class=\"katex--display\"><span class=\"katex-display\"><span class=\"katex\"><span class=\"katex-mathml\"><\/p>\n<p>         y<\/p>\n<p>         &#061;<\/p>\n<p>         f<\/p>\n<p>         (<\/p>\n<p>          w<\/p>\n<p>          T<\/p>\n<p>         x<\/p>\n<p>         &#043;<\/p>\n<p>         b<\/p>\n<p>         )<\/p>\n<p>         y &#061; f(w^T x &#043; b) <\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.625em;vertical-align: -0.1944em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0359em\">y<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 1.1413em;vertical-align: -0.25em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.1076em\">f<\/span><span class=\"mopen\">(<\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.0269em\">w<\/span><span class=\"msupsub\"><span class=\"vlist-t\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.8913em\"><span class=\"\" style=\"top: -3.113em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mathnormal mtight\" style=\"margin-right: 0.1389em\">T<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mord mathnormal\">x<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">&#043;<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 1em;vertical-align: -0.25em\"><\/span><span class=\"mord mathnormal\">b<\/span><span class=\"mclose\">)<\/span><\/span><\/span><\/span><\/span><\/span><\/p>\n<p>\u5176\u4e2d&#xff1a;<\/p>\n<ul>\n<li><span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">\n<p>         x<\/p>\n<p>        x<\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.4306em\"><\/span><span class=\"mord mathnormal\">x<\/span><\/span><\/span><\/span><\/span>&#xff1a;\u8f93\u5165\u5411\u91cf&#xff0c;<span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\"><\/p>\n<p>         x<\/p>\n<p>         &#061;<\/p>\n<p>         [<\/p>\n<p>          x<\/p>\n<p>          1<\/p>\n<p>         ,<\/p>\n<p>          x<\/p>\n<p>          2<\/p>\n<p>         ,<\/p>\n<p>         .<\/p>\n<p>         .<\/p>\n<p>         .<\/p>\n<p>         ,<\/p>\n<p>          x<\/p>\n<p>          n<\/p>\n<p>          ]<\/p>\n<p>          T<\/p>\n<p>        x &#061; [x_1, x_2, &#8230;, x_n]^T<\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.4306em\"><\/span><span class=\"mord mathnormal\">x<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 1.0913em;vertical-align: -0.25em\"><\/span><span class=\"mopen\">[<\/span><span class=\"mord\"><span class=\"mord mathnormal\">x<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.3011em\"><span class=\"\" style=\"top: -2.55em;margin-left: 0em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\">1<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mpunct\">,<\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\">x<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.3011em\"><span class=\"\" style=\"top: -2.55em;margin-left: 0em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\">2<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mpunct\">,<\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord\">&#8230;<\/span><span class=\"mpunct\">,<\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\">x<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.1514em\"><span class=\"\" style=\"top: -2.55em;margin-left: 0em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mathnormal mtight\">n<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mclose\"><span class=\"mclose\">]<\/span><span class=\"msupsub\"><span class=\"vlist-t\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.8413em\"><span class=\"\" style=\"top: -3.063em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mathnormal mtight\" style=\"margin-right: 0.1389em\">T<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/li>\n<li><span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">\n<p>         w<\/p>\n<p>        w<\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.4306em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0269em\">w<\/span><\/span><\/span><\/span><\/span>&#xff1a;\u6743\u91cd\u5411\u91cf&#xff0c;<span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\"><\/p>\n<p>         w<\/p>\n<p>         &#061;<\/p>\n<p>         [<\/p>\n<p>          w<\/p>\n<p>          1<\/p>\n<p>         ,<\/p>\n<p>          w<\/p>\n<p>          2<\/p>\n<p>         ,<\/p>\n<p>         .<\/p>\n<p>         .<\/p>\n<p>         .<\/p>\n<p>         ,<\/p>\n<p>          w<\/p>\n<p>          n<\/p>\n<p>          ]<\/p>\n<p>          T<\/p>\n<p>        w &#061; [w_1, w_2, &#8230;, w_n]^T<\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.4306em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0269em\">w<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 1.0913em;vertical-align: -0.25em\"><\/span><span class=\"mopen\">[<\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.0269em\">w<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.3011em\"><span class=\"\" style=\"top: -2.55em;margin-left: -0.0269em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\">1<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mpunct\">,<\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.0269em\">w<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.3011em\"><span class=\"\" style=\"top: -2.55em;margin-left: -0.0269em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\">2<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mpunct\">,<\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord\">&#8230;<\/span><span class=\"mpunct\">,<\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.0269em\">w<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.1514em\"><span class=\"\" style=\"top: -2.55em;margin-left: -0.0269em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mathnormal mtight\">n<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mclose\"><span class=\"mclose\">]<\/span><span class=\"msupsub\"><span class=\"vlist-t\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.8413em\"><span class=\"\" style=\"top: -3.063em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mathnormal mtight\" style=\"margin-right: 0.1389em\">T<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/li>\n<li><span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">\n<p>         b<\/p>\n<p>        b<\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.6944em\"><\/span><span class=\"mord mathnormal\">b<\/span><\/span><\/span><\/span><\/span>&#xff1a;\u504f\u7f6e\u9879<\/li>\n<li><span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">\n<p>         f<\/p>\n<p>         (<\/p>\n<p>         \u22c5<\/p>\n<p>         )<\/p>\n<p>        f(\\\\cdot)<\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 1em;vertical-align: -0.25em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.1076em\">f<\/span><span class=\"mopen\">(<\/span><span class=\"mord\">\u22c5<\/span><span class=\"mclose\">)<\/span><\/span><\/span><\/span><\/span>&#xff1a;\u6fc0\u6d3b\u51fd\u6570<\/li>\n<li><span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">\n<p>         y<\/p>\n<p>        y<\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.625em;vertical-align: -0.1944em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0359em\">y<\/span><\/span><\/span><\/span><\/span>&#xff1a;\u8f93\u51fa<\/li>\n<\/ul>\n<h5>1.1.3 \u6fc0\u6d3b\u51fd\u6570&#xff1a;\u5f15\u5165\u975e\u7ebf\u6027\u7684\u5173\u952e<\/h5>\n<p>\u5e38\u89c1\u7684\u6fc0\u6d3b\u51fd\u6570&#xff1a;<\/p>\n<li>\n<p>Sigmoid\u51fd\u6570&#xff1a; <span class=\"katex--display\"><span class=\"katex-display\"><span class=\"katex\"><span class=\"katex-mathml\"><\/p>\n<p>           f<\/p>\n<p>           (<\/p>\n<p>           z<\/p>\n<p>           )<\/p>\n<p>           &#061;<\/p>\n<p>            1<\/p>\n<p>             1<\/p>\n<p>             &#043;<\/p>\n<p>              e<\/p>\n<p>               \u2212<\/p>\n<p>               z<\/p>\n<p>           f(z) &#061; \\\\frac{1}{1 &#043; e^{-z}} <\/p>\n<p>       <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 1em;vertical-align: -0.25em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.1076em\">f<\/span><span class=\"mopen\">(<\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.044em\">z<\/span><span class=\"mclose\">)<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 2.0908em;vertical-align: -0.7693em\"><\/span><span class=\"mord\"><span class=\"mopen nulldelimiter\"><\/span><span class=\"mfrac\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.3214em\"><span class=\"\" style=\"top: -2.314em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"mord\"><span class=\"mord\">1<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">&#043;<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\">e<\/span><span class=\"msupsub\"><span class=\"vlist-t\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.6973em\"><span class=\"\" style=\"top: -2.989em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mtight\">\u2212<\/span><span class=\"mord mathnormal mtight\" style=\"margin-right: 0.044em\">z<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"\" style=\"top: -3.23em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"frac-line\" style=\"border-bottom-width: 0.04em\"><\/span><\/span><span class=\"\" style=\"top: -3.677em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"mord\"><span class=\"mord\">1<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.7693em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><span class=\"mclose nulldelimiter\"><\/span><\/span><\/span><\/span><\/span><\/span><\/span> \u5bfc\u6570&#xff1a;<span class=\"katex--display\"><span class=\"katex-display\"><span class=\"katex\"><span class=\"katex-mathml\"><\/p>\n<p>            f<\/p>\n<p>            \u2032<\/p>\n<p>           (<\/p>\n<p>           z<\/p>\n<p>           )<\/p>\n<p>           &#061;<\/p>\n<p>           f<\/p>\n<p>           (<\/p>\n<p>           z<\/p>\n<p>           )<\/p>\n<p>           (<\/p>\n<p>           1<\/p>\n<p>           \u2212<\/p>\n<p>           f<\/p>\n<p>           (<\/p>\n<p>           z<\/p>\n<p>           )<\/p>\n<p>           )<\/p>\n<p>           f&#039;(z) &#061; f(z)(1 &#8211; f(z)) <\/p>\n<p>       <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 1.0519em;vertical-align: -0.25em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.1076em\">f<\/span><span class=\"msupsub\"><span class=\"vlist-t\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.8019em\"><span class=\"\" style=\"top: -3.113em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mtight\">\u2032<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mopen\">(<\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.044em\">z<\/span><span class=\"mclose\">)<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 1em;vertical-align: -0.25em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.1076em\">f<\/span><span class=\"mopen\">(<\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.044em\">z<\/span><span class=\"mclose\">)<\/span><span class=\"mopen\">(<\/span><span class=\"mord\">1<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u2212<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 1em;vertical-align: -0.25em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.1076em\">f<\/span><span class=\"mopen\">(<\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.044em\">z<\/span><span class=\"mclose\">))<\/span><\/span><\/span><\/span><\/span><\/span><\/p>\n<\/li>\n<li>\n<p>ReLU\u51fd\u6570&#xff08;Rectified Linear Unit&#xff09;&#xff1a; <span class=\"katex--display\"><span class=\"katex-display\"><span class=\"katex\"><span class=\"katex-mathml\"><\/p>\n<p>           f<\/p>\n<p>           (<\/p>\n<p>           z<\/p>\n<p>           )<\/p>\n<p>           &#061;<\/p>\n<p>           m<\/p>\n<p>           a<\/p>\n<p>           x<\/p>\n<p>           (<\/p>\n<p>           0<\/p>\n<p>           ,<\/p>\n<p>           z<\/p>\n<p>           )<\/p>\n<p>           f(z) &#061; max(0, z) <\/p>\n<p>       <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 1em;vertical-align: -0.25em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.1076em\">f<\/span><span class=\"mopen\">(<\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.044em\">z<\/span><span class=\"mclose\">)<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 1em;vertical-align: -0.25em\"><\/span><span class=\"mord mathnormal\">ma<\/span><span class=\"mord mathnormal\">x<\/span><span class=\"mopen\">(<\/span><span class=\"mord\">0<\/span><span class=\"mpunct\">,<\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.044em\">z<\/span><span class=\"mclose\">)<\/span><\/span><\/span><\/span><\/span><\/span> \u5bfc\u6570&#xff1a;<span class=\"katex--display\"><span class=\"katex-display\"><span class=\"katex\"><span class=\"katex-mathml\"><\/p>\n<p>            f<\/p>\n<p>            \u2032<\/p>\n<p>           (<\/p>\n<p>           z<\/p>\n<p>           )<\/p>\n<p>           &#061;<\/p>\n<p>            {<\/p>\n<p>                 1<\/p>\n<p>                 ,<\/p>\n<p>                 z<\/p>\n<p>                 &gt;<\/p>\n<p>                 0<\/p>\n<p>                 0<\/p>\n<p>                 ,<\/p>\n<p>                 z<\/p>\n<p>                 \u2264<\/p>\n<p>                 0<\/p>\n<p>           f&#039;(z) &#061; \\\\begin{cases} 1, &amp; z &gt; 0 \\\\\\\\ 0, &amp; z \\\\leq 0 \\\\end{cases} <\/p>\n<p>       <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 1.0519em;vertical-align: -0.25em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.1076em\">f<\/span><span class=\"msupsub\"><span class=\"vlist-t\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.8019em\"><span class=\"\" style=\"top: -3.113em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mtight\">\u2032<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mopen\">(<\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.044em\">z<\/span><span class=\"mclose\">)<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 3em;vertical-align: -1.25em\"><\/span><span class=\"minner\"><span class=\"mopen delimcenter\" style=\"top: 0em\"><span class=\"delimsizing size4\">{<\/span><\/span><span class=\"mord\"><span class=\"mtable\"><span class=\"col-align-l\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.69em\"><span class=\"\" style=\"top: -3.69em\"><span class=\"pstrut\" style=\"height: 3.008em\"><\/span><span class=\"mord\"><span class=\"mord\">1<\/span><span class=\"mpunct\">,<\/span><\/span><\/span><span class=\"\" style=\"top: -2.25em\"><span class=\"pstrut\" style=\"height: 3.008em\"><\/span><span class=\"mord\"><span class=\"mord\">0<\/span><span class=\"mpunct\">,<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.19em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><span class=\"arraycolsep\" style=\"width: 1em\"><\/span><span class=\"col-align-l\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.69em\"><span class=\"\" style=\"top: -3.69em\"><span class=\"pstrut\" style=\"height: 3.008em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.044em\">z<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&gt;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mord\">0<\/span><\/span><\/span><span class=\"\" style=\"top: -2.25em\"><span class=\"pstrut\" style=\"height: 3.008em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.044em\">z<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">\u2264<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mord\">0<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.19em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mclose nulldelimiter\"><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/p>\n<\/li>\n<li>\n<p>Tanh\u51fd\u6570&#xff1a; <span class=\"katex--display\"><span class=\"katex-display\"><span class=\"katex\"><span class=\"katex-mathml\"><\/p>\n<p>           f<\/p>\n<p>           (<\/p>\n<p>           z<\/p>\n<p>           )<\/p>\n<p>           &#061;<\/p>\n<p>              e<\/p>\n<p>              z<\/p>\n<p>             \u2212<\/p>\n<p>              e<\/p>\n<p>               \u2212<\/p>\n<p>               z<\/p>\n<p>              e<\/p>\n<p>              z<\/p>\n<p>             &#043;<\/p>\n<p>              e<\/p>\n<p>               \u2212<\/p>\n<p>               z<\/p>\n<p>           f(z) &#061; \\\\frac{e^z &#8211; e^{-z}}{e^z &#043; e^{-z}} <\/p>\n<p>       <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 1em;vertical-align: -0.25em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.1076em\">f<\/span><span class=\"mopen\">(<\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.044em\">z<\/span><span class=\"mclose\">)<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 2.2177em;vertical-align: -0.7693em\"><\/span><span class=\"mord\"><span class=\"mopen nulldelimiter\"><\/span><span class=\"mfrac\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.4483em\"><span class=\"\" style=\"top: -2.314em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"mord\"><span class=\"mord\"><span class=\"mord mathnormal\">e<\/span><span class=\"msupsub\"><span class=\"vlist-t\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.5904em\"><span class=\"\" style=\"top: -2.989em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mathnormal mtight\" style=\"margin-right: 0.044em\">z<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">&#043;<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\">e<\/span><span class=\"msupsub\"><span class=\"vlist-t\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.6973em\"><span class=\"\" style=\"top: -2.989em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mtight\">\u2212<\/span><span class=\"mord mathnormal mtight\" style=\"margin-right: 0.044em\">z<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"\" style=\"top: -3.23em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"frac-line\" style=\"border-bottom-width: 0.04em\"><\/span><\/span><span class=\"\" style=\"top: -3.677em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"mord\"><span class=\"mord\"><span class=\"mord mathnormal\">e<\/span><span class=\"msupsub\"><span class=\"vlist-t\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.6644em\"><span class=\"\" style=\"top: -3.063em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mathnormal mtight\" style=\"margin-right: 0.044em\">z<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u2212<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\">e<\/span><span class=\"msupsub\"><span class=\"vlist-t\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.7713em\"><span class=\"\" style=\"top: -3.063em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mtight\">\u2212<\/span><span class=\"mord mathnormal mtight\" style=\"margin-right: 0.044em\">z<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.7693em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><span class=\"mclose nulldelimiter\"><\/span><\/span><\/span><\/span><\/span><\/span><\/span> \u5bfc\u6570&#xff1a;<span class=\"katex--display\"><span class=\"katex-display\"><span class=\"katex\"><span class=\"katex-mathml\"><\/p>\n<p>            f<\/p>\n<p>            \u2032<\/p>\n<p>           (<\/p>\n<p>           z<\/p>\n<p>           )<\/p>\n<p>           &#061;<\/p>\n<p>           1<\/p>\n<p>           \u2212<\/p>\n<p>           f<\/p>\n<p>           (<\/p>\n<p>           z<\/p>\n<p>            )<\/p>\n<p>            2<\/p>\n<p>           f&#039;(z) &#061; 1 &#8211; f(z)^2 <\/p>\n<p>       <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 1.0519em;vertical-align: -0.25em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.1076em\">f<\/span><span class=\"msupsub\"><span class=\"vlist-t\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.8019em\"><span class=\"\" style=\"top: -3.113em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mtight\">\u2032<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mopen\">(<\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.044em\">z<\/span><span class=\"mclose\">)<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.7278em;vertical-align: -0.0833em\"><\/span><span class=\"mord\">1<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u2212<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 1.1141em;vertical-align: -0.25em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.1076em\">f<\/span><span class=\"mopen\">(<\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.044em\">z<\/span><span class=\"mclose\"><span class=\"mclose\">)<\/span><span class=\"msupsub\"><span class=\"vlist-t\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.8641em\"><span class=\"\" style=\"top: -3.113em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\">2<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/p>\n<\/li>\n<h5>1.1.4 \u795e\u7ecf\u7f51\u7edc\u7684\u57fa\u672c\u7ed3\u6784<\/h5>\n<p>ANN\u901a\u5e38\u7531\u8f93\u5165\u5c42\u3001\u9690\u85cf\u5c42\u548c\u8f93\u51fa\u5c42\u7ec4\u6210&#xff1a;<\/p>\n<ul>\n<li>\u8f93\u5165\u5c42&#xff1a;\u63a5\u6536\u5916\u90e8\u8f93\u5165\u6570\u636e<\/li>\n<li>\u9690\u85cf\u5c42&#xff1a;\u5b66\u4e60\u8f93\u5165\u6570\u636e\u7684\u7279\u5f81\u8868\u793a<\/li>\n<li>\u8f93\u51fa\u5c42&#xff1a;\u4ea7\u751f\u6700\u7ec8\u7684\u9884\u6d4b\u7ed3\u679c<\/li>\n<\/ul>\n<h5>1.1.5 \u53cd\u5411\u4f20\u64ad\u7b97\u6cd5&#xff1a;\u795e\u7ecf\u7f51\u7edc\u7684\u5b66\u4e60\u65b9\u5f0f<\/h5>\n<p>\u53cd\u5411\u4f20\u64ad\u7b97\u6cd5&#xff08;Backpropagation&#xff09;\u662f\u8bad\u7ec3ANN\u7684\u6838\u5fc3\u7b97\u6cd5&#xff0c;\u57fa\u4e8e\u68af\u5ea6\u4e0b\u964d\u6cd5&#xff1a;<\/p>\n<li>\u524d\u5411\u4f20\u64ad&#xff1a;\u8ba1\u7b97\u7f51\u7edc\u8f93\u51fa<\/li>\n<li>\u8ba1\u7b97\u635f\u5931&#xff1a;\u4f7f\u7528\u635f\u5931\u51fd\u6570\u8ba1\u7b97\u9884\u6d4b\u503c\u4e0e\u771f\u5b9e\u503c\u7684\u5dee\u5f02<\/li>\n<li>\u53cd\u5411\u4f20\u64ad&#xff1a;\u8ba1\u7b97\u635f\u5931\u5bf9\u5404\u5c42\u53c2\u6570\u7684\u68af\u5ea6<\/li>\n<li>\u53c2\u6570\u66f4\u65b0&#xff1a;\u4f7f\u7528\u68af\u5ea6\u4e0b\u964d\u6cd5\u66f4\u65b0\u7f51\u7edc\u53c2\u6570<\/li>\n<p>\u635f\u5931\u51fd\u6570&#xff08;\u4ea4\u53c9\u71b5\u635f\u5931&#xff09;&#xff1a; <span class=\"katex--display\"><span class=\"katex-display\"><span class=\"katex\"><span class=\"katex-mathml\"><\/p>\n<p>         L<\/p>\n<p>         &#061;<\/p>\n<p>         \u2212<\/p>\n<p>          \u2211<\/p>\n<p>           i<\/p>\n<p>           &#061;<\/p>\n<p>           1<\/p>\n<p>          C<\/p>\n<p>          y<\/p>\n<p>          i<\/p>\n<p>         log<\/p>\n<p>         \u2061<\/p>\n<p>         (<\/p>\n<p>           y<\/p>\n<p>           ^<\/p>\n<p>          i<\/p>\n<p>         )<\/p>\n<p>         L &#061; -\\\\sum_{i&#061;1}^{C} y_i \\\\log(\\\\hat{y}_i) <\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.6833em\"><\/span><span class=\"mord mathnormal\">L<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 3.106em;vertical-align: -1.2777em\"><\/span><span class=\"mord\">\u2212<\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mop op-limits\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.8283em\"><span class=\"\" style=\"top: -1.8723em;margin-left: 0em\"><span class=\"pstrut\" style=\"height: 3.05em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mathnormal mtight\">i<\/span><span class=\"mrel mtight\">&#061;<\/span><span class=\"mord mtight\">1<\/span><\/span><\/span><\/span><span class=\"\" style=\"top: -3.05em\"><span class=\"pstrut\" style=\"height: 3.05em\"><\/span><span class=\"\"><span class=\"mop op-symbol large-op\">\u2211<\/span><\/span><\/span><span class=\"\" style=\"top: -4.3em;margin-left: 0em\"><span class=\"pstrut\" style=\"height: 3.05em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mathnormal mtight\" style=\"margin-right: 0.0715em\">C<\/span><\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.2777em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.0359em\">y<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.3117em\"><span class=\"\" style=\"top: -2.55em;margin-left: -0.0359em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mathnormal mtight\">i<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mop\">lo<span style=\"margin-right: 0.0139em\">g<\/span><\/span><span class=\"mopen\">(<\/span><span class=\"mord\"><span class=\"mord accent\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.6944em\"><span class=\"\" style=\"top: -3em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0359em\">y<\/span><\/span><span class=\"\" style=\"top: -3em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"accent-body\" style=\"left: -0.1944em\"><span class=\"mord\">^<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.1944em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.3117em\"><span class=\"\" style=\"top: -2.55em;margin-left: -0.0359em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mathnormal mtight\">i<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mclose\">)<\/span><\/span><\/span><\/span><\/span><\/span><\/p>\n<p>\u5176\u4e2d&#xff1a;<\/p>\n<ul>\n<li><span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">\n<p>         C<\/p>\n<p>        C<\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.6833em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0715em\">C<\/span><\/span><\/span><\/span><\/span>&#xff1a;\u7c7b\u522b\u6570<\/li>\n<li><span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">\n<p>          y<\/p>\n<p>          i<\/p>\n<p>        y_i<\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.625em;vertical-align: -0.1944em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.0359em\">y<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.3117em\"><span class=\"\" style=\"top: -2.55em;margin-left: -0.0359em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mathnormal mtight\">i<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span>&#xff1a;\u771f\u5b9e\u6807\u7b7e\u7684\u72ec\u70ed\u7f16\u7801<\/li>\n<li><span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">\n<p>           y<\/p>\n<p>           ^<\/p>\n<p>          i<\/p>\n<p>        \\\\hat{y}_i<\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.8889em;vertical-align: -0.1944em\"><\/span><span class=\"mord\"><span class=\"mord accent\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.6944em\"><span class=\"\" style=\"top: -3em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0359em\">y<\/span><\/span><span class=\"\" style=\"top: -3em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"accent-body\" style=\"left: -0.1944em\"><span class=\"mord\">^<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.1944em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.3117em\"><span class=\"\" style=\"top: -2.55em;margin-left: -0.0359em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mathnormal mtight\">i<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span>&#xff1a;\u7f51\u7edc\u9884\u6d4b\u7684\u6982\u7387<\/li>\n<\/ul>\n<p>\u53c2\u6570\u66f4\u65b0&#xff1a; <span class=\"katex--display\"><span class=\"katex-display\"><span class=\"katex\"><span class=\"katex-mathml\"><\/p>\n<p>         w<\/p>\n<p>         &#061;<\/p>\n<p>         w<\/p>\n<p>         \u2212<\/p>\n<p>         \u03b7<\/p>\n<p>           \u2202<\/p>\n<p>           L<\/p>\n<p>           \u2202<\/p>\n<p>           w<\/p>\n<p>         w &#061; w &#8211; \\\\eta \\\\frac{\\\\partial L}{\\\\partial w} <\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.4306em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0269em\">w<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.6667em;vertical-align: -0.0833em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0269em\">w<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u2212<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 2.0574em;vertical-align: -0.686em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0359em\">\u03b7<\/span><span class=\"mord\"><span class=\"mopen nulldelimiter\"><\/span><span class=\"mfrac\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.3714em\"><span class=\"\" style=\"top: -2.314em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"mord\"><span class=\"mord\" style=\"margin-right: 0.0556em\">\u2202<\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0269em\">w<\/span><\/span><\/span><span class=\"\" style=\"top: -3.23em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"frac-line\" style=\"border-bottom-width: 0.04em\"><\/span><\/span><span class=\"\" style=\"top: -3.677em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"mord\"><span class=\"mord\" style=\"margin-right: 0.0556em\">\u2202<\/span><span class=\"mord mathnormal\">L<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.686em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><span class=\"mclose nulldelimiter\"><\/span><\/span><\/span><\/span><\/span><\/span><\/span> <span class=\"katex--display\"><span class=\"katex-display\"><span class=\"katex\"><span class=\"katex-mathml\"><\/p>\n<p>         b<\/p>\n<p>         &#061;<\/p>\n<p>         b<\/p>\n<p>         \u2212<\/p>\n<p>         \u03b7<\/p>\n<p>           \u2202<\/p>\n<p>           L<\/p>\n<p>           \u2202<\/p>\n<p>           b<\/p>\n<p>         b &#061; b &#8211; \\\\eta \\\\frac{\\\\partial L}{\\\\partial b} <\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.6944em\"><\/span><span class=\"mord mathnormal\">b<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.7778em;vertical-align: -0.0833em\"><\/span><span class=\"mord mathnormal\">b<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u2212<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 2.0574em;vertical-align: -0.686em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0359em\">\u03b7<\/span><span class=\"mord\"><span class=\"mopen nulldelimiter\"><\/span><span class=\"mfrac\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.3714em\"><span class=\"\" style=\"top: -2.314em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"mord\"><span class=\"mord\" style=\"margin-right: 0.0556em\">\u2202<\/span><span class=\"mord mathnormal\">b<\/span><\/span><\/span><span class=\"\" style=\"top: -3.23em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"frac-line\" style=\"border-bottom-width: 0.04em\"><\/span><\/span><span class=\"\" style=\"top: -3.677em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"mord\"><span class=\"mord\" style=\"margin-right: 0.0556em\">\u2202<\/span><span class=\"mord mathnormal\">L<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.686em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><span class=\"mclose nulldelimiter\"><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/p>\n<p>\u5176\u4e2d&#xff1a;<\/p>\n<ul>\n<li><span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">\n<p>         \u03b7<\/p>\n<p>        \\\\eta<\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.625em;vertical-align: -0.1944em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0359em\">\u03b7<\/span><\/span><\/span><\/span><\/span>&#xff1a;\u5b66\u4e60\u7387<\/li>\n<li><span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">\n<p>           \u2202<\/p>\n<p>           L<\/p>\n<p>           \u2202<\/p>\n<p>           w<\/p>\n<p>        \\\\frac{\\\\partial L}{\\\\partial w}<\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 1.2251em;vertical-align: -0.345em\"><\/span><span class=\"mord\"><span class=\"mopen nulldelimiter\"><\/span><span class=\"mfrac\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.8801em\"><span class=\"\" style=\"top: -2.655em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mtight\" style=\"margin-right: 0.0556em\">\u2202<\/span><span class=\"mord mathnormal mtight\" style=\"margin-right: 0.0269em\">w<\/span><\/span><\/span><\/span><span class=\"\" style=\"top: -3.23em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"frac-line\" style=\"border-bottom-width: 0.04em\"><\/span><\/span><span class=\"\" style=\"top: -3.394em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mtight\" style=\"margin-right: 0.0556em\">\u2202<\/span><span class=\"mord mathnormal mtight\">L<\/span><\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.345em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><span class=\"mclose nulldelimiter\"><\/span><\/span><\/span><\/span><\/span><\/span>&#xff1a;\u635f\u5931\u5bf9\u6743\u91cd\u7684\u68af\u5ea6<\/li>\n<li><span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">\n<p>           \u2202<\/p>\n<p>           L<\/p>\n<p>           \u2202<\/p>\n<p>           b<\/p>\n<p>        \\\\frac{\\\\partial L}{\\\\partial b}<\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 1.2251em;vertical-align: -0.345em\"><\/span><span class=\"mord\"><span class=\"mopen nulldelimiter\"><\/span><span class=\"mfrac\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.8801em\"><span class=\"\" style=\"top: -2.655em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mtight\" style=\"margin-right: 0.0556em\">\u2202<\/span><span class=\"mord mathnormal mtight\">b<\/span><\/span><\/span><\/span><span class=\"\" style=\"top: -3.23em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"frac-line\" style=\"border-bottom-width: 0.04em\"><\/span><\/span><span class=\"\" style=\"top: -3.394em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mtight\" style=\"margin-right: 0.0556em\">\u2202<\/span><span class=\"mord mathnormal mtight\">L<\/span><\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.345em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><span class=\"mclose nulldelimiter\"><\/span><\/span><\/span><\/span><\/span><\/span>&#xff1a;\u635f\u5931\u5bf9\u504f\u7f6e\u7684\u68af\u5ea6<\/li>\n<\/ul>\n<h4>1.2 \u6df1\u5ea6\u795e\u7ecf\u7f51\u7edc&#xff08;DNN&#xff09;&#xff1a;\u63a2\u7d22\u66f4\u6df1\u5c42\u6b21\u7684\u7279\u5f81<\/h4>\n<h5>1.2.1 \u6df1\u5ea6\u7f51\u7edc\u7684\u5b9a\u4e49\u4e0e\u7279\u70b9<\/h5>\n<p>\u6df1\u5ea6\u795e\u7ecf\u7f51\u7edc&#xff08;Deep Neural Network, DNN&#xff09;\u662f\u5177\u6709\u591a\u4e2a\u9690\u85cf\u5c42\u7684\u795e\u7ecf\u7f51\u7edc&#xff0c;\u80fd\u591f\u5b66\u4e60\u66f4\u590d\u6742\u7684\u7279\u5f81\u8868\u793a\u3002<\/p>\n<h5>1.2.2 \u6df1\u5ea6\u7f51\u7edc\u7684\u6570\u5b66\u8868\u8fbe<\/h5>\n<p>DNN\u7684\u7ed3\u6784\u4e0eANN\u7c7b\u4f3c&#xff0c;\u4f46\u5305\u542b\u66f4\u591a\u7684\u9690\u85cf\u5c42&#xff1a;<\/p>\n<p><span class=\"katex--display\"><span class=\"katex-display\"><span class=\"katex\"><span class=\"katex-mathml\"><\/p>\n<p>         y<\/p>\n<p>         &#061;<\/p>\n<p>          f<\/p>\n<p>          L<\/p>\n<p>         (<\/p>\n<p>          f<\/p>\n<p>           L<\/p>\n<p>           \u2212<\/p>\n<p>           1<\/p>\n<p>         (<\/p>\n<p>         .<\/p>\n<p>         .<\/p>\n<p>         .<\/p>\n<p>          f<\/p>\n<p>          2<\/p>\n<p>         (<\/p>\n<p>          f<\/p>\n<p>          1<\/p>\n<p>         (<\/p>\n<p>          w<\/p>\n<p>          1<\/p>\n<p>          T<\/p>\n<p>         x<\/p>\n<p>         &#043;<\/p>\n<p>          b<\/p>\n<p>          1<\/p>\n<p>         )<\/p>\n<p>          w<\/p>\n<p>          2<\/p>\n<p>          T<\/p>\n<p>         &#043;<\/p>\n<p>          b<\/p>\n<p>          2<\/p>\n<p>         )<\/p>\n<p>         .<\/p>\n<p>         .<\/p>\n<p>         .<\/p>\n<p>         )<\/p>\n<p>          w<\/p>\n<p>          L<\/p>\n<p>          T<\/p>\n<p>         &#043;<\/p>\n<p>          b<\/p>\n<p>          L<\/p>\n<p>         )<\/p>\n<p>         y &#061; f_L( f_{L-1}( &#8230; f_2( f_1(w_1^T x &#043; b_1) w_2^T &#043; b_2 ) &#8230; ) w_L^T &#043; b_L ) <\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.625em;vertical-align: -0.1944em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0359em\">y<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 1.1413em;vertical-align: -0.25em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.1076em\">f<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.3283em\"><span class=\"\" style=\"top: -2.55em;margin-left: -0.1076em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mathnormal mtight\">L<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mopen\">(<\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.1076em\">f<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.3283em\"><span class=\"\" style=\"top: -2.55em;margin-left: -0.1076em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mathnormal mtight\">L<\/span><span class=\"mbin mtight\">\u2212<\/span><span class=\"mord mtight\">1<\/span><\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.2083em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mopen\">(<\/span><span class=\"mord\">&#8230;<\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.1076em\">f<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.3011em\"><span class=\"\" style=\"top: -2.55em;margin-left: -0.1076em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\">2<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mopen\">(<\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.1076em\">f<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.3011em\"><span class=\"\" style=\"top: -2.55em;margin-left: -0.1076em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\">1<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mopen\">(<\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.0269em\">w<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.8913em\"><span class=\"\" style=\"top: -2.453em;margin-left: -0.0269em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\">1<\/span><\/span><\/span><span class=\"\" style=\"top: -3.113em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mathnormal mtight\" style=\"margin-right: 0.1389em\">T<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.247em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mord mathnormal\">x<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">&#043;<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 1.1413em;vertical-align: -0.25em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\">b<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.3011em\"><span class=\"\" style=\"top: -2.55em;margin-left: 0em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\">1<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mclose\">)<\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.0269em\">w<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.8913em\"><span class=\"\" style=\"top: -2.453em;margin-left: -0.0269em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\">2<\/span><\/span><\/span><span class=\"\" style=\"top: -3.113em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mathnormal mtight\" style=\"margin-right: 0.1389em\">T<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.247em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">&#043;<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 1.1413em;vertical-align: -0.25em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\">b<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.3011em\"><span class=\"\" style=\"top: -2.55em;margin-left: 0em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\">2<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mclose\">)<\/span><span class=\"mord\">&#8230;<\/span><span class=\"mclose\">)<\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.0269em\">w<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.8913em\"><span class=\"\" style=\"top: -2.453em;margin-left: -0.0269em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mathnormal mtight\">L<\/span><\/span><\/span><span class=\"\" style=\"top: -3.113em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mathnormal mtight\" style=\"margin-right: 0.1389em\">T<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.247em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">&#043;<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 1em;vertical-align: -0.25em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\">b<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.3283em\"><span class=\"\" style=\"top: -2.55em;margin-left: 0em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mathnormal mtight\">L<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mclose\">)<\/span><\/span><\/span><\/span><\/span><\/span><\/p>\n<p>\u5176\u4e2d&#xff1a;<\/p>\n<ul>\n<li><span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">\n<p>         L<\/p>\n<p>        L<\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.6833em\"><\/span><span class=\"mord mathnormal\">L<\/span><\/span><\/span><\/span><\/span>&#xff1a;\u7f51\u7edc\u5c42\u6570<\/li>\n<li><span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">\n<p>          f<\/p>\n<p>          l<\/p>\n<p>         (<\/p>\n<p>         \u22c5<\/p>\n<p>         )<\/p>\n<p>        f_l(\\\\cdot)<\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 1em;vertical-align: -0.25em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.1076em\">f<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.3361em\"><span class=\"\" style=\"top: -2.55em;margin-left: -0.1076em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mathnormal mtight\" style=\"margin-right: 0.0197em\">l<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mopen\">(<\/span><span class=\"mord\">\u22c5<\/span><span class=\"mclose\">)<\/span><\/span><\/span><\/span><\/span>&#xff1a;\u7b2c<span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\"><\/p>\n<p>         l<\/p>\n<p>        l<\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.6944em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0197em\">l<\/span><\/span><\/span><\/span><\/span>\u5c42\u7684\u6fc0\u6d3b\u51fd\u6570<\/li>\n<li><span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">\n<p>          w<\/p>\n<p>          l<\/p>\n<p>        w_l<\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.5806em;vertical-align: -0.15em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.0269em\">w<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.3361em\"><span class=\"\" style=\"top: -2.55em;margin-left: -0.0269em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mathnormal mtight\" style=\"margin-right: 0.0197em\">l<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span>&#xff1a;\u7b2c<span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\"><\/p>\n<p>         l<\/p>\n<p>        l<\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.6944em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0197em\">l<\/span><\/span><\/span><\/span><\/span>\u5c42\u7684\u6743\u91cd\u77e9\u9635<\/li>\n<li><span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">\n<p>          b<\/p>\n<p>          l<\/p>\n<p>        b_l<\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.8444em;vertical-align: -0.15em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\">b<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.3361em\"><span class=\"\" style=\"top: -2.55em;margin-left: 0em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mathnormal mtight\" style=\"margin-right: 0.0197em\">l<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span>&#xff1a;\u7b2c<span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\"><\/p>\n<p>         l<\/p>\n<p>        l<\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.6944em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0197em\">l<\/span><\/span><\/span><\/span><\/span>\u5c42\u7684\u504f\u7f6e\u5411\u91cf<\/li>\n<\/ul>\n<h5>1.2.3 \u6df1\u5ea6\u7f51\u7edc\u7684\u8bad\u7ec3\u6311\u6218\u4e0e\u89e3\u51b3\u65b9\u6848<\/h5>\n<p>\u6df1\u5ea6\u7f51\u7edc\u8bad\u7ec3\u9762\u4e34\u7684\u6311\u6218&#xff1a;<\/p>\n<li>\n<p>\u68af\u5ea6\u6d88\u5931\/\u7206\u70b8&#xff1a;\u6df1\u5ea6\u7f51\u7edc\u4e2d&#xff0c;\u68af\u5ea6\u5728\u53cd\u5411\u4f20\u64ad\u8fc7\u7a0b\u4e2d\u5bb9\u6613\u6d88\u5931\u6216\u7206\u70b8<\/p>\n<ul>\n<li>\u89e3\u51b3\u65b9\u6cd5&#xff1a;\u4f7f\u7528ReLU\u6fc0\u6d3b\u51fd\u6570\u3001\u6279\u6807\u51c6\u5316\u3001\u6b8b\u5dee\u8fde\u63a5\u7b49<\/li>\n<\/ul>\n<\/li>\n<li>\n<p>\u8fc7\u62df\u5408&#xff1a;\u6df1\u5ea6\u7f51\u7edc\u5bb9\u6613\u5728\u8bad\u7ec3\u6570\u636e\u4e0a\u8fc7\u5ea6\u62df\u5408<\/p>\n<ul>\n<li>\u89e3\u51b3\u65b9\u6cd5&#xff1a;\u6b63\u5219\u5316\u3001Dropout\u3001\u6570\u636e\u589e\u5f3a\u7b49<\/li>\n<\/ul>\n<\/li>\n<li>\n<p>\u8ba1\u7b97\u590d\u6742\u5ea6\u9ad8&#xff1a;\u6df1\u5ea6\u7f51\u7edc\u9700\u8981\u5927\u91cf\u7684\u8ba1\u7b97\u8d44\u6e90\u548c\u8bad\u7ec3\u6570\u636e<\/p>\n<\/li>\n<h4>1.3 \u5377\u79ef\u795e\u7ecf\u7f51\u7edc&#xff08;CNN&#xff09;<\/h4>\n<h5>1.3.1 CNN\u7684\u57fa\u672c\u601d\u60f3<\/h5>\n<p>\u5377\u79ef\u795e\u7ecf\u7f51\u7edc&#xff08;Convolutional Neural Network, CNN&#xff09;\u662f\u4e00\u79cd\u4e13\u95e8\u7528\u4e8e\u5904\u7406\u5177\u6709\u7f51\u683c\u7ed3\u6784\u6570\u636e&#xff08;\u5982\u56fe\u50cf\u3001\u97f3\u9891&#xff09;\u7684\u795e\u7ecf\u7f51\u7edc&#xff0c;\u901a\u8fc7\u5377\u79ef\u64cd\u4f5c\u63d0\u53d6\u5c40\u90e8\u7279\u5f81\u3002<\/p>\n<h5>1.3.2 CNN\u7684\u6838\u5fc3\u7ec4\u4ef6\u89e3\u6790<\/h5>\n<li>\n<p>\u5377\u79ef\u5c42<\/p>\n<ul>\n<li>\u5377\u79ef\u64cd\u4f5c&#xff1a; <span class=\"katex--display\"><span class=\"katex-display\"><span class=\"katex\"><span class=\"katex-mathml\">\n<p>            (<\/p>\n<p>            f<\/p>\n<p>            \u2217<\/p>\n<p>            g<\/p>\n<p>            )<\/p>\n<p>            (<\/p>\n<p>            i<\/p>\n<p>            ,<\/p>\n<p>            j<\/p>\n<p>            )<\/p>\n<p>            &#061;<\/p>\n<p>             \u2211<\/p>\n<p>              m<\/p>\n<p>              &#061;<\/p>\n<p>              \u2212<\/p>\n<p>              M<\/p>\n<p>             M<\/p>\n<p>             \u2211<\/p>\n<p>              n<\/p>\n<p>              &#061;<\/p>\n<p>              \u2212<\/p>\n<p>              N<\/p>\n<p>             N<\/p>\n<p>            f<\/p>\n<p>            (<\/p>\n<p>            m<\/p>\n<p>            ,<\/p>\n<p>            n<\/p>\n<p>            )<\/p>\n<p>            g<\/p>\n<p>            (<\/p>\n<p>            i<\/p>\n<p>            &#043;<\/p>\n<p>            m<\/p>\n<p>            ,<\/p>\n<p>            j<\/p>\n<p>            &#043;<\/p>\n<p>            n<\/p>\n<p>            )<\/p>\n<p>            (f * g)(i, j) &#061; \\\\sum_{m&#061;-M}^{M} \\\\sum_{n&#061;-N}^{N} f(m, n) g(i&#043;m, j&#043;n) <\/p>\n<p>        <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 1em;vertical-align: -0.25em\"><\/span><span class=\"mopen\">(<\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.1076em\">f<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u2217<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 1em;vertical-align: -0.25em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0359em\">g<\/span><span class=\"mclose\">)<\/span><span class=\"mopen\">(<\/span><span class=\"mord mathnormal\">i<\/span><span class=\"mpunct\">,<\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0572em\">j<\/span><span class=\"mclose\">)<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 3.181em;vertical-align: -1.3527em\"><\/span><span class=\"mop op-limits\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.8283em\"><span class=\"\" style=\"top: -1.8557em;margin-left: 0em\"><span class=\"pstrut\" style=\"height: 3.05em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mathnormal mtight\">m<\/span><span class=\"mrel mtight\">&#061;<\/span><span class=\"mord mtight\">\u2212<\/span><span class=\"mord mathnormal mtight\" style=\"margin-right: 0.109em\">M<\/span><\/span><\/span><\/span><span class=\"\" style=\"top: -3.05em\"><span class=\"pstrut\" style=\"height: 3.05em\"><\/span><span class=\"\"><span class=\"mop op-symbol large-op\">\u2211<\/span><\/span><\/span><span class=\"\" style=\"top: -4.3em;margin-left: 0em\"><span class=\"pstrut\" style=\"height: 3.05em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mathnormal mtight\" style=\"margin-right: 0.109em\">M<\/span><\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.3527em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mop op-limits\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.8283em\"><span class=\"\" style=\"top: -1.8557em;margin-left: 0em\"><span class=\"pstrut\" style=\"height: 3.05em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mathnormal mtight\">n<\/span><span class=\"mrel mtight\">&#061;<\/span><span class=\"mord mtight\">\u2212<\/span><span class=\"mord mathnormal mtight\" style=\"margin-right: 0.109em\">N<\/span><\/span><\/span><\/span><span class=\"\" style=\"top: -3.05em\"><span class=\"pstrut\" style=\"height: 3.05em\"><\/span><span class=\"\"><span class=\"mop op-symbol large-op\">\u2211<\/span><\/span><\/span><span class=\"\" style=\"top: -4.3em;margin-left: 0em\"><span class=\"pstrut\" style=\"height: 3.05em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mathnormal mtight\" style=\"margin-right: 0.109em\">N<\/span><\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.3527em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.1076em\">f<\/span><span class=\"mopen\">(<\/span><span class=\"mord mathnormal\">m<\/span><span class=\"mpunct\">,<\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord mathnormal\">n<\/span><span class=\"mclose\">)<\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0359em\">g<\/span><span class=\"mopen\">(<\/span><span class=\"mord mathnormal\">i<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">&#043;<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.854em;vertical-align: -0.1944em\"><\/span><span class=\"mord mathnormal\">m<\/span><span class=\"mpunct\">,<\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0572em\">j<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">&#043;<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 1em;vertical-align: -0.25em\"><\/span><span class=\"mord mathnormal\">n<\/span><span class=\"mclose\">)<\/span><\/span><\/span><\/span><\/span><\/span> \u5176\u4e2d&#xff0c;<span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\"><\/p>\n<p>           f<\/p>\n<p>          f<\/p>\n<p>       <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.8889em;vertical-align: -0.1944em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.1076em\">f<\/span><\/span><\/span><\/span><\/span>\u662f\u5377\u79ef\u6838&#xff0c;<span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\"><\/p>\n<p>           g<\/p>\n<p>          g<\/p>\n<p>       <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.625em;vertical-align: -0.1944em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0359em\">g<\/span><\/span><\/span><\/span><\/span>\u662f\u8f93\u5165\u7279\u5f81\u56fe<\/li>\n<li>\u53c2\u6570\u5171\u4eab&#xff1a;\u540c\u4e00\u5377\u79ef\u6838\u5728\u6574\u4e2a\u8f93\u5165\u4e0a\u5171\u4eab\u53c2\u6570<\/li>\n<li>\u5c40\u90e8\u8fde\u63a5&#xff1a;\u6bcf\u4e2a\u795e\u7ecf\u5143\u53ea\u8fde\u63a5\u8f93\u5165\u7684\u5c40\u90e8\u533a\u57df<\/li>\n<\/ul>\n<\/li>\n<li>\n<p>\u6c60\u5316\u5c42<\/p>\n<ul>\n<li>\u6700\u5927\u6c60\u5316&#xff08;Max Pooling&#xff09;&#xff1a; <span class=\"katex--display\"><span class=\"katex-display\"><span class=\"katex\"><span class=\"katex-mathml\">\n<p>            (<\/p>\n<p>             P<\/p>\n<p>              m<\/p>\n<p>              a<\/p>\n<p>              x<\/p>\n<p>            g<\/p>\n<p>            )<\/p>\n<p>            (<\/p>\n<p>            i<\/p>\n<p>            ,<\/p>\n<p>            j<\/p>\n<p>            )<\/p>\n<p>            &#061;<\/p>\n<p>            m<\/p>\n<p>            a<\/p>\n<p>             x<\/p>\n<p>              m<\/p>\n<p>              &#061;<\/p>\n<p>              0<\/p>\n<p>              k<\/p>\n<p>              \u2212<\/p>\n<p>              1<\/p>\n<p>            m<\/p>\n<p>            a<\/p>\n<p>             x<\/p>\n<p>              n<\/p>\n<p>              &#061;<\/p>\n<p>              0<\/p>\n<p>              k<\/p>\n<p>              \u2212<\/p>\n<p>              1<\/p>\n<p>            g<\/p>\n<p>            (<\/p>\n<p>            i<\/p>\n<p>            \u2217<\/p>\n<p>            k<\/p>\n<p>            &#043;<\/p>\n<p>            m<\/p>\n<p>            ,<\/p>\n<p>            j<\/p>\n<p>            \u2217<\/p>\n<p>            k<\/p>\n<p>            &#043;<\/p>\n<p>            n<\/p>\n<p>            )<\/p>\n<p>            (P_{max} g)(i, j) &#061; max_{m&#061;0}^{k-1} max_{n&#061;0}^{k-1} g(i*k&#043;m, j*k&#043;n) <\/p>\n<p>        <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 1em;vertical-align: -0.25em\"><\/span><span class=\"mopen\">(<\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.1389em\">P<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.1514em\"><span class=\"\" style=\"top: -2.55em;margin-left: -0.1389em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mathnormal mtight\">ma<\/span><span class=\"mord mathnormal mtight\">x<\/span><\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0359em\">g<\/span><span class=\"mclose\">)<\/span><span class=\"mopen\">(<\/span><span class=\"mord mathnormal\">i<\/span><span class=\"mpunct\">,<\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0572em\">j<\/span><span class=\"mclose\">)<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 1.1555em;vertical-align: -0.2564em\"><\/span><span class=\"mord mathnormal\">ma<\/span><span class=\"mord\"><span class=\"mord mathnormal\">x<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.8991em\"><span class=\"\" style=\"top: -2.4436em;margin-left: 0em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mathnormal mtight\">m<\/span><span class=\"mrel mtight\">&#061;<\/span><span class=\"mord mtight\">0<\/span><\/span><\/span><\/span><span class=\"\" style=\"top: -3.113em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mathnormal mtight\" style=\"margin-right: 0.0315em\">k<\/span><span class=\"mbin mtight\">\u2212<\/span><span class=\"mord mtight\">1<\/span><\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.2564em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mord mathnormal\">ma<\/span><span class=\"mord\"><span class=\"mord mathnormal\">x<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.8991em\"><span class=\"\" style=\"top: -2.4436em;margin-left: 0em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mathnormal mtight\">n<\/span><span class=\"mrel mtight\">&#061;<\/span><span class=\"mord mtight\">0<\/span><\/span><\/span><\/span><span class=\"\" style=\"top: -3.113em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mathnormal mtight\" style=\"margin-right: 0.0315em\">k<\/span><span class=\"mbin mtight\">\u2212<\/span><span class=\"mord mtight\">1<\/span><\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.2564em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0359em\">g<\/span><span class=\"mopen\">(<\/span><span class=\"mord mathnormal\">i<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u2217<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.7778em;vertical-align: -0.0833em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0315em\">k<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">&#043;<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.854em;vertical-align: -0.1944em\"><\/span><span class=\"mord mathnormal\">m<\/span><span class=\"mpunct\">,<\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0572em\">j<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u2217<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.7778em;vertical-align: -0.0833em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0315em\">k<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">&#043;<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 1em;vertical-align: -0.25em\"><\/span><span class=\"mord mathnormal\">n<\/span><span class=\"mclose\">)<\/span><\/span><\/span><\/span><\/span><\/span><\/li>\n<li>\u5e73\u5747\u6c60\u5316&#xff08;Average Pooling&#xff09;&#xff1a; <span class=\"katex--display\"><span class=\"katex-display\"><span class=\"katex\"><span class=\"katex-mathml\">\n<p>            (<\/p>\n<p>             P<\/p>\n<p>              a<\/p>\n<p>              v<\/p>\n<p>              g<\/p>\n<p>            g<\/p>\n<p>            )<\/p>\n<p>            (<\/p>\n<p>            i<\/p>\n<p>            ,<\/p>\n<p>            j<\/p>\n<p>            )<\/p>\n<p>            &#061;<\/p>\n<p>             1<\/p>\n<p>              k<\/p>\n<p>              2<\/p>\n<p>             \u2211<\/p>\n<p>              m<\/p>\n<p>              &#061;<\/p>\n<p>              0<\/p>\n<p>              k<\/p>\n<p>              \u2212<\/p>\n<p>              1<\/p>\n<p>             \u2211<\/p>\n<p>              n<\/p>\n<p>              &#061;<\/p>\n<p>              0<\/p>\n<p>              k<\/p>\n<p>              \u2212<\/p>\n<p>              1<\/p>\n<p>            g<\/p>\n<p>            (<\/p>\n<p>            i<\/p>\n<p>            \u2217<\/p>\n<p>            k<\/p>\n<p>            &#043;<\/p>\n<p>            m<\/p>\n<p>            ,<\/p>\n<p>            j<\/p>\n<p>            \u2217<\/p>\n<p>            k<\/p>\n<p>            &#043;<\/p>\n<p>            n<\/p>\n<p>            )<\/p>\n<p>            (P_{avg} g)(i, j) &#061; \\\\frac{1}{k^2} \\\\sum_{m&#061;0}^{k-1} \\\\sum_{n&#061;0}^{k-1} g(i*k&#043;m, j*k&#043;n) <\/p>\n<p>        <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 1.0361em;vertical-align: -0.2861em\"><\/span><span class=\"mopen\">(<\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.1389em\">P<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.1514em\"><span class=\"\" style=\"top: -2.55em;margin-left: -0.1389em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mathnormal mtight\">a<\/span><span class=\"mord mathnormal mtight\" style=\"margin-right: 0.0359em\">vg<\/span><\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.2861em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0359em\">g<\/span><span class=\"mclose\">)<\/span><span class=\"mopen\">(<\/span><span class=\"mord mathnormal\">i<\/span><span class=\"mpunct\">,<\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0572em\">j<\/span><span class=\"mclose\">)<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 3.1032em;vertical-align: -1.2671em\"><\/span><span class=\"mord\"><span class=\"mopen nulldelimiter\"><\/span><span class=\"mfrac\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.3214em\"><span class=\"\" style=\"top: -2.314em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"mord\"><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.0315em\">k<\/span><span class=\"msupsub\"><span class=\"vlist-t\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.7401em\"><span class=\"\" style=\"top: -2.989em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\">2<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"\" style=\"top: -3.23em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"frac-line\" style=\"border-bottom-width: 0.04em\"><\/span><\/span><span class=\"\" style=\"top: -3.677em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"mord\"><span class=\"mord\">1<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.686em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><span class=\"mclose nulldelimiter\"><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mop op-limits\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.8361em\"><span class=\"\" style=\"top: -1.8829em;margin-left: 0em\"><span class=\"pstrut\" style=\"height: 3.05em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mathnormal mtight\">m<\/span><span class=\"mrel mtight\">&#061;<\/span><span class=\"mord mtight\">0<\/span><\/span><\/span><\/span><span class=\"\" style=\"top: -3.05em\"><span class=\"pstrut\" style=\"height: 3.05em\"><\/span><span class=\"\"><span class=\"mop op-symbol large-op\">\u2211<\/span><\/span><\/span><span class=\"\" style=\"top: -4.3em;margin-left: 0em\"><span class=\"pstrut\" style=\"height: 3.05em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mathnormal mtight\" style=\"margin-right: 0.0315em\">k<\/span><span class=\"mbin mtight\">\u2212<\/span><span class=\"mord mtight\">1<\/span><\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.2671em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mop op-limits\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.8361em\"><span class=\"\" style=\"top: -1.8829em;margin-left: 0em\"><span class=\"pstrut\" style=\"height: 3.05em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mathnormal mtight\">n<\/span><span class=\"mrel mtight\">&#061;<\/span><span class=\"mord mtight\">0<\/span><\/span><\/span><\/span><span class=\"\" style=\"top: -3.05em\"><span class=\"pstrut\" style=\"height: 3.05em\"><\/span><span class=\"\"><span class=\"mop op-symbol large-op\">\u2211<\/span><\/span><\/span><span class=\"\" style=\"top: -4.3em;margin-left: 0em\"><span class=\"pstrut\" style=\"height: 3.05em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mathnormal mtight\" style=\"margin-right: 0.0315em\">k<\/span><span class=\"mbin mtight\">\u2212<\/span><span class=\"mord mtight\">1<\/span><\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.2671em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0359em\">g<\/span><span class=\"mopen\">(<\/span><span class=\"mord mathnormal\">i<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u2217<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.7778em;vertical-align: -0.0833em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0315em\">k<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">&#043;<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.854em;vertical-align: -0.1944em\"><\/span><span class=\"mord mathnormal\">m<\/span><span class=\"mpunct\">,<\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0572em\">j<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u2217<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.7778em;vertical-align: -0.0833em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0315em\">k<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">&#043;<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 1em;vertical-align: -0.25em\"><\/span><span class=\"mord mathnormal\">n<\/span><span class=\"mclose\">)<\/span><\/span><\/span><\/span><\/span><\/span><\/li>\n<li>\u4f5c\u7528&#xff1a;\u964d\u7ef4\u3001\u63d0\u53d6\u4e3b\u8981\u7279\u5f81\u3001\u589e\u5f3a\u4e0d\u53d8\u6027<\/li>\n<\/ul>\n<\/li>\n<li>\n<p>\u5168\u8fde\u63a5\u5c42&#xff1a;\u5c06\u5377\u79ef\u5c42\u63d0\u53d6\u7684\u7279\u5f81\u6620\u5c04\u5230\u8f93\u51fa\u7c7b\u522b<\/p>\n<\/li>\n<li>\n<p>\u6fc0\u6d3b\u51fd\u6570&#xff1a;\u901a\u5e38\u4f7f\u7528ReLU<\/p>\n<\/li>\n<h5>1.3.3 \u7ecf\u5178CNN\u67b6\u6784\u4e00\u89c8<\/h5>\n<ul>\n<li>LeNet-5&#xff1a;\u6700\u65e9\u7684CNN\u67b6\u6784\u4e4b\u4e00&#xff0c;\u7528\u4e8e\u624b\u5199\u6570\u5b57\u8bc6\u522b<\/li>\n<li>AlexNet&#xff1a;2012\u5e74ImageNet\u51a0\u519b&#xff0c;\u5f15\u5165\u4e86ReLU\u3001Dropout\u3001\u6570\u636e\u589e\u5f3a\u7b49\u6280\u672f<\/li>\n<li>VGGNet&#xff1a;\u4f7f\u7528\u66f4\u5c0f\u7684\u5377\u79ef\u6838\u548c\u66f4\u6df1\u7684\u7f51\u7edc\u7ed3\u6784<\/li>\n<li>ResNet&#xff1a;\u5f15\u5165\u6b8b\u5dee\u8fde\u63a5&#xff0c;\u89e3\u51b3\u6df1\u5ea6\u7f51\u7edc\u8bad\u7ec3\u95ee\u9898<\/li>\n<li>Inception&#xff1a;\u4f7f\u7528\u591a\u5c3a\u5ea6\u5377\u79ef\u6838\u5e76\u884c\u5904\u7406<\/li>\n<\/ul>\n<h3>\u4e8c\u3001\u6df1\u5ea6\u5b66\u4e60\u6a21\u578b\u5bf9\u6bd4&#xff1a;ANN\u3001CNN\u4e0eDNN\u7684\u4f18\u52a3\u52bf<\/h3>\n<h4>2.1 \u4e09\u8005\u7684\u5173\u7cfb&#xff1a;\u4ece\u57fa\u7840\u5230\u4e13\u4e1a<\/h4>\n<ul>\n<li>ANN&#xff1a;\u662f\u6240\u6709\u795e\u7ecf\u7f51\u7edc\u7684\u57fa\u7840\u6846\u67b6&#xff0c;\u5305\u62ec\u4e86\u6240\u6709\u5177\u6709\u795e\u7ecf\u5143\u7ed3\u6784\u7684\u7f51\u7edc<\/li>\n<li>DNN&#xff1a;\u662fANN\u7684\u6df1\u5ea6\u7248\u672c&#xff0c;\u901a\u8fc7\u589e\u52a0\u9690\u85cf\u5c42\u6df1\u5ea6\u63d0\u5347\u8868\u8fbe\u80fd\u529b<\/li>\n<li>CNN&#xff1a;\u662fDNN\u7684\u7279\u6b8a\u53d8\u4f53&#xff0c;\u9488\u5bf9\u7f51\u683c\u7ed3\u6784\u6570\u636e\u4f18\u5316\u8bbe\u8ba1<\/li>\n<\/ul>\n<h4>2.2 \u6838\u5fc3\u7279\u6027\u5bf9\u6bd4<\/h4>\n<table>\n<tr>\u7279\u6027ANNDNNCNN<\/tr>\n<tbody>\n<tr>\n<td>\u7ed3\u6784<\/td>\n<td>\u5355\u9690\u85cf\u5c42\u6216\u591a\u9690\u85cf\u5c42<\/td>\n<td>\u591a\u5c42\u9690\u85cf\u5c42<\/td>\n<td>\u5377\u79ef\u5c42&#043;\u6c60\u5316\u5c42&#043;\u5168\u8fde\u63a5\u5c42<\/td>\n<\/tr>\n<tr>\n<td>\u6570\u636e\u7c7b\u578b<\/td>\n<td>\u901a\u7528<\/td>\n<td>\u901a\u7528<\/td>\n<td>\u7f51\u683c\u7ed3\u6784&#xff08;\u56fe\u50cf\u3001\u97f3\u9891\u7b49&#xff09;<\/td>\n<\/tr>\n<tr>\n<td>\u53c2\u6570\u6570\u91cf<\/td>\n<td>\u8f83\u5c11<\/td>\n<td>\u8f83\u591a<\/td>\n<td>\u76f8\u5bf9\u8f83\u5c11&#xff08;\u53c2\u6570\u5171\u4eab&#xff09;<\/td>\n<\/tr>\n<tr>\n<td>\u7279\u5f81\u63d0\u53d6<\/td>\n<td>\u4eba\u5de5\u6216\u81ea\u52a8<\/td>\n<td>\u81ea\u52a8<\/td>\n<td>\u81ea\u52a8&#xff08;\u5c40\u90e8\u7279\u5f81&#xff09;<\/td>\n<\/tr>\n<tr>\n<td>\u4e3b\u8981\u5e94\u7528<\/td>\n<td>\u7b80\u5355\u5206\u7c7b\u3001\u56de\u5f52<\/td>\n<td>\u590d\u6742\u5206\u7c7b\u3001\u56de\u5f52<\/td>\n<td>\u56fe\u50cf\u8bc6\u522b\u3001\u8bed\u97f3\u8bc6\u522b<\/td>\n<\/tr>\n<tr>\n<td>\u8bad\u7ec3\u96be\u5ea6<\/td>\n<td>\u8f83\u4f4e<\/td>\n<td>\u8f83\u9ad8&#xff08;\u68af\u5ea6\u6d88\u5931\/\u7206\u70b8&#xff09;<\/td>\n<td>\u4e2d\u7b49&#xff08;\u5377\u79ef\u64cd\u4f5c\u7b80\u5316\u8bad\u7ec3&#xff09;<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<h4>2.3 \u9002\u7528\u573a\u666f\u5206\u6790<\/h4>\n<li>\n<p>ANN&#xff1a;<\/p>\n<ul>\n<li>\u9002\u7528\u4e8e\u7b80\u5355\u7684\u6a21\u5f0f\u8bc6\u522b\u4efb\u52a1<\/li>\n<li>\u8f93\u5165\u6570\u636e\u7ef4\u5ea6\u8f83\u4f4e\u7684\u60c5\u51b5<\/li>\n<li>\u8d44\u6e90\u53d7\u9650\u7684\u73af\u5883<\/li>\n<\/ul>\n<\/li>\n<li>\n<p>DNN&#xff1a;<\/p>\n<ul>\n<li>\u9002\u7528\u4e8e\u590d\u6742\u7684\u975e\u7ebf\u6027\u95ee\u9898<\/li>\n<li>\u9700\u8981\u5b66\u4e60\u6df1\u5c42\u7279\u5f81\u8868\u793a\u7684\u4efb\u52a1<\/li>\n<li>\u6709\u5927\u91cf\u8bad\u7ec3\u6570\u636e\u7684\u60c5\u51b5<\/li>\n<\/ul>\n<\/li>\n<li>\n<p>CNN&#xff1a;<\/p>\n<ul>\n<li>\u56fe\u50cf\u5904\u7406&#xff1a;\u56fe\u50cf\u5206\u7c7b\u3001\u76ee\u6807\u68c0\u6d4b\u3001\u56fe\u50cf\u5206\u5272<\/li>\n<li>\u97f3\u9891\u5904\u7406&#xff1a;\u8bed\u97f3\u8bc6\u522b\u3001\u97f3\u9891\u5206\u7c7b<\/li>\n<li>\u89c6\u9891\u5904\u7406&#xff1a;\u89c6\u9891\u5206\u7c7b\u3001\u52a8\u4f5c\u8bc6\u522b<\/li>\n<li>\u5176\u4ed6\u7f51\u683c\u7ed3\u6784\u6570\u636e&#xff1a;\u6587\u672c\u5904\u7406&#xff08;1D CNN&#xff09;\u3001\u65f6\u95f4\u5e8f\u5217\u5206\u6790<\/li>\n<\/ul>\n<\/li>\n<h3>\u4e09\u3001\u9879\u76ee\u5b9e\u8df5&#xff1a;CNN\u4e0eDNN\u7ed3\u5408\u7684\u56fe\u50cf\u5206\u7c7b\u65b9\u6848<\/h3>\n<h4>3.1 \u9879\u76ee\u80cc\u666f\u4e0e\u76ee\u6807<\/h4>\n<p>\u672c\u9879\u76ee\u65e8\u5728\u4f7f\u7528CNN\u4e0eDNN\u7ed3\u5408\u7684\u67b6\u6784&#xff0c;\u5b9e\u73b0CIFAR10\u6570\u636e\u96c6\u7684\u9ad8\u6548\u56fe\u50cf\u5206\u7c7b&#xff0c;\u63a2\u7d22\u4e24\u79cd\u7f51\u7edc\u7ed3\u6784\u7684\u4f18\u52bf\u4e92\u8865\u3002<\/p>\n<h4>3.2 \u4e3a\u4ec0\u4e48\u9009\u62e9CNN&#043;DNN\u7684\u7ec4\u5408&#xff1f;<\/h4>\n<li>\n<p>\u4f18\u52bf\u4e92\u8865&#xff1a;<\/p>\n<ul>\n<li>CNN&#xff1a;\u64c5\u957f\u63d0\u53d6\u56fe\u50cf\u7684\u5c40\u90e8\u7279\u5f81&#xff0c;\u4fdd\u7559\u7a7a\u95f4\u4fe1\u606f&#xff0c;\u53c2\u6570\u5171\u4eab\u51cf\u5c11\u8ba1\u7b97\u91cf<\/li>\n<li>DNN&#xff1a;\u64c5\u957f\u5bf9\u63d0\u53d6\u7684\u7279\u5f81\u8fdb\u884c\u590d\u6742\u7684\u975e\u7ebf\u6027\u53d8\u6362\u548c\u5206\u7c7b<\/li>\n<\/ul>\n<\/li>\n<li>\n<p>\u7279\u5f81\u5c42\u6b21\u5316&#xff1a;<\/p>\n<ul>\n<li>CNN\u5c42\u4ece\u5e95\u5c42\u5230\u9ad8\u5c42\u63d0\u53d6\u8d8a\u6765\u8d8a\u62bd\u8c61\u7684\u7279\u5f81<\/li>\n<li>DNN\u5c42\u5bf9\u8fd9\u4e9b\u5c42\u6b21\u5316\u7684\u7279\u5f81\u8fdb\u884c\u7efc\u5408\u548c\u5206\u7c7b<\/li>\n<\/ul>\n<\/li>\n<li>\n<p>\u63d0\u9ad8\u6cdb\u5316\u80fd\u529b&#xff1a;<\/p>\n<ul>\n<li>CNN\u7684\u5c40\u90e8\u8fde\u63a5\u548c\u53c2\u6570\u5171\u4eab\u51cf\u5c11\u4e86\u8fc7\u62df\u5408\u98ce\u9669<\/li>\n<li>DNN\u7684\u6df1\u5ea6\u7ed3\u6784\u80fd\u591f\u5b66\u4e60\u66f4\u590d\u6742\u7684\u6a21\u5f0f<\/li>\n<\/ul>\n<\/li>\n<li>\n<p>\u9ad8\u6548\u5904\u7406\u56fe\u50cf\u6570\u636e&#xff1a;<\/p>\n<ul>\n<li>CNN\u4e13\u95e8\u9488\u5bf9\u56fe\u50cf\u6570\u636e\u7684\u7ed3\u6784\u8bbe\u8ba1<\/li>\n<li>\u907f\u514d\u4e86\u5168\u8fde\u63a5\u7f51\u7edc\u5904\u7406\u56fe\u50cf\u65f6\u7684\u53c2\u6570\u7206\u70b8\u95ee\u9898<\/li>\n<\/ul>\n<\/li>\n<h3>\u56db\u3001\u4ee3\u7801\u5b9e\u73b0\u4e0e\u8bbe\u8ba1\u601d\u8def\u89e3\u6790<\/h3>\n<h4>4.1 \u9879\u76ee\u4ee3\u7801\u7ed3\u6784\u6982\u89c8<\/h4>\n<p>CNN_DNN.py<br \/>\n\u251c\u2500\u2500 \u9879\u76ee\u8bf4\u660e\u6587\u6863<br \/>\n\u251c\u2500\u2500 \u6838\u5fc3\u5e93\u5bfc\u5165<br \/>\n\u251c\u2500\u2500 \u6570\u636e\u9884\u5904\u7406\u6a21\u5757<br \/>\n\u2502   \u251c\u2500\u2500 \u6570\u636e\u589e\u5f3a\u7b56\u7565<br \/>\n\u2502   \u251c\u2500\u2500 \u6570\u636e\u5f52\u4e00\u5316\u5904\u7406<br \/>\n\u2502   \u251c\u2500\u2500 \u6570\u636e\u96c6\u52a0\u8f7d\u4e0e\u62c6\u5206<br \/>\n\u251c\u2500\u2500 \u6a21\u578b\u5b9a\u4e49&#xff08;CNN_DNN_Model\u7c7b&#xff09;<br \/>\n\u2502   \u251c\u2500\u2500 CNN\u7279\u5f81\u63d0\u53d6\u6a21\u5757<br \/>\n\u2502   \u251c\u2500\u2500 DNN\u5206\u7c7b\u51b3\u7b56\u6a21\u5757<br \/>\n\u2502   \u251c\u2500\u2500 \u524d\u5411\u4f20\u64ad\u6d41\u7a0b<br \/>\n\u251c\u2500\u2500 \u8f85\u52a9\u5de5\u5177\u51fd\u6570<br \/>\n\u2502   \u251c\u2500\u2500 \u8bad\u7ec3\u8fdb\u5ea6\u6761<br \/>\n\u251c\u2500\u2500 \u6838\u5fc3\u529f\u80fd\u51fd\u6570<br \/>\n\u2502   \u251c\u2500\u2500 \u6a21\u578b\u8bad\u7ec3\u51fd\u6570<br \/>\n\u2502   \u251c\u2500\u2500 \u6a21\u578b\u6d4b\u8bd5\u51fd\u6570<br \/>\n\u251c\u2500\u2500 \u53ef\u89c6\u5316\u6a21\u5757<br \/>\n\u2502   \u251c\u2500\u2500 \u8bad\u7ec3\u635f\u5931\u66f2\u7ebf\u7ed8\u5236<br \/>\n\u2502   \u251c\u2500\u2500 \u6df7\u6dc6\u77e9\u9635\u751f\u6210<br \/>\n\u2502   \u251c\u2500\u2500 \u9519\u8bef\u5206\u7c7b\u6837\u672c\u5c55\u793a<br \/>\n\u2502   \u251c\u2500\u2500 \u6b63\u786e\u5206\u7c7b\u6837\u672c\u5c55\u793a<br \/>\n\u2502   \u251c\u2500\u2500 \u9884\u6d4b\u6982\u7387\u5206\u5e03\u53ef\u89c6\u5316<br \/>\n\u2514\u2500\u2500 \u4e3b\u7a0b\u5e8f\u5165\u53e3<\/p>\n<h4>4.2 \u6a21\u578b\u8bbe\u8ba1\u4e0e\u5b9e\u73b0\u601d\u8def<\/h4>\n<h5>4.2.1 \u7f51\u7edc\u67b6\u6784\u8bbe\u8ba1&#xff1a;\u4ece\u7279\u5f81\u63d0\u53d6\u5230\u5206\u7c7b\u51b3\u7b56<\/h5>\n<h6>CNN\u7279\u5f81\u63d0\u53d6\u90e8\u5206&#xff1a;\u6355\u6349\u56fe\u50cf\u5c40\u90e8\u4fe1\u606f<\/h6>\n<ul>\n<li>\n<p>3\u4e2a\u5377\u79ef\u5757\u7684\u8bbe\u8ba1\u4f9d\u636e&#xff1a;<\/p>\n<ul>\n<li>CIFAR10\u56fe\u50cf\u5c3a\u5bf8\u4e3a32\u00d732\u00d73&#xff0c;\u5c5e\u4e8e\u5c0f\u5c3a\u5bf8\u56fe\u50cf&#xff0c;3\u4e2a\u5377\u79ef\u5757\u8db3\u4ee5\u63d0\u53d6\u4ece\u4f4e\u7ea7\u5230\u9ad8\u7ea7\u7684\u7279\u5f81<\/li>\n<li>\u7b2c\u4e00\u5c42&#xff08;\u4f4e\u7ea7\u7279\u5f81&#xff09;&#xff1a;\u63d0\u53d6\u8fb9\u7f18\u3001\u7eb9\u7406\u7b49\u57fa\u7840\u7279\u5f81<\/li>\n<li>\u7b2c\u4e8c\u5c42&#xff08;\u4e2d\u7ea7\u7279\u5f81&#xff09;&#xff1a;\u7ec4\u5408\u4f4e\u7ea7\u7279\u5f81&#xff0c;\u5f62\u6210\u5f62\u72b6\u3001\u8f6e\u5ed3\u7b49<\/li>\n<li>\u7b2c\u4e09\u5c42&#xff08;\u9ad8\u7ea7\u7279\u5f81&#xff09;&#xff1a;\u5b66\u4e60\u66f4\u590d\u6742\u7684\u8bed\u4e49\u7279\u5f81&#xff0c;\u5982\u7269\u4f53\u90e8\u4ef6<\/li>\n<\/ul>\n<\/li>\n<li>\n<p>\u6bcf\u4e2a\u5377\u79ef\u5757\u7684\u7ed3\u6784&#xff1a;<\/p>\n<p> \u5377\u79ef\u5757 &#061; \u5377\u79ef\u5c42 &#043; ReLU\u6fc0\u6d3b\u51fd\u6570 &#043; \u6700\u5927\u6c60\u5316\u5c42<\/p>\n<ul>\n<li>\u5377\u79ef\u5c42&#xff1a;\u4f7f\u75283\u00d73\u5377\u79ef\u6838&#xff0c;\u8fd9\u662fCNN\u4e2d\u6700\u6709\u6548\u7684\u5377\u79ef\u6838\u5c3a\u5bf8&#xff0c;\u5728\u4fdd\u6301\u8f83\u5c0f\u611f\u53d7\u91ce\u7684\u540c\u65f6\u5e73\u8861\u4e86\u8ba1\u7b97\u6548\u7387<\/li>\n<li>\u901a\u9053\u6570\u8bbe\u8ba1&#xff1a;3\u219232\u219264\u2192128&#xff0c;\u5448\u6307\u6570\u589e\u957f&#xff0c;\u7b26\u5408\u7279\u5f81\u8868\u793a\u7684\u5c42\u6b21\u5316\u9700\u6c42<\/li>\n<li>\u6c60\u5316\u5c42&#xff1a;\u6bcf\u5c42\u4f7f\u75282\u00d72\u6700\u5927\u6c60\u5316&#xff0c;\u6b65\u957f2&#xff0c;\u4f7f\u7279\u5f81\u56fe\u5c3a\u5bf8\u51cf\u534a<\/li>\n<\/ul>\n<\/li>\n<li>\n<p>\u7ef4\u5ea6\u6f14\u5316\u8fc7\u7a0b&#xff1a;<\/p>\n<p> \u8f93\u5165&#xff1a;32\u00d732\u00d73<br \/>\n\u5377\u79ef\u57571&#xff1a;32\u00d732\u00d73 \u2192 32\u00d732\u00d732 \u2192 16\u00d716\u00d732<br \/>\n\u5377\u79ef\u57572&#xff1a;16\u00d716\u00d732 \u2192 16\u00d716\u00d764 \u2192 8\u00d78\u00d764<br \/>\n\u5377\u79ef\u57573&#xff1a;8\u00d78\u00d764 \u2192 8\u00d78\u00d7128 \u2192 4\u00d74\u00d7128\n <\/li>\n<\/ul>\n<h6>DNN\u5206\u7c7b\u51b3\u7b56\u90e8\u5206&#xff1a;\u5b66\u4e60\u590d\u6742\u975e\u7ebf\u6027\u6620\u5c04<\/h6>\n<ul>\n<li>\n<p>4\u4e2a\u5168\u8fde\u63a5\u5c42\u7684\u8bbe\u8ba1\u4f9d\u636e&#xff1a;<\/p>\n<ul>\n<li>\u591a\u5c42\u5168\u8fde\u63a5\u5c42\u53ef\u4ee5\u5b66\u4e60\u590d\u6742\u7684\u975e\u7ebf\u6027\u6620\u5c04\u5173\u7cfb<\/li>\n<li>\u91c7\u7528&#034;\u6f0f\u6597&#034;\u5f62\u72b6&#xff08;2048\u2192512\u2192256\u2192128\u219210&#xff09;&#xff0c;\u9010\u6b65\u51cf\u5c11\u7279\u5f81\u7ef4\u5ea6&#xff0c;\u538b\u7f29\u4fe1\u606f<\/li>\n<li>\u6700\u540e\u4e00\u5c42\u4e3a10\u4e2a\u795e\u7ecf\u5143&#xff0c;\u5bf9\u5e94CIFAR10\u768410\u4e2a\u7c7b\u522b<\/li>\n<\/ul>\n<\/li>\n<li>\n<p>Dropout\u5c42\u7684\u8bbe\u8ba1&#xff1a;<\/p>\n<ul>\n<li>\u5728\u6bcf\u4e2a\u5168\u8fde\u63a5\u5c42\u540e\u6dfb\u52a0Dropout\u5c42&#xff0c;dropout\u7387\u4e3a0.2<\/li>\n<li>\u8f83\u5c0f\u7684dropout\u7387\u5728\u9632\u6b62\u8fc7\u62df\u5408\u7684\u540c\u65f6\u4fdd\u7559\u4e86\u5927\u90e8\u5206\u7279\u5f81\u4fe1\u606f<\/li>\n<li>\u4ec5\u5728\u8bad\u7ec3\u65f6\u4f7f\u7528&#xff0c;\u6d4b\u8bd5\u65f6\u5173\u95ed<\/li>\n<\/ul>\n<\/li>\n<\/ul>\n<h6>\u7ef4\u5ea6\u5339\u914d\u8bbe\u8ba1<\/h6>\n<ul>\n<li>\n<p>\u5377\u79ef\u5c42\u8f93\u51fa\u8ba1\u7b97&#xff1a;<\/p>\n<p> <span class=\"token comment\"># \u5377\u79ef\u5c42\u8f93\u51fa\u5c3a\u5bf8\u8ba1\u7b97<\/span><br \/>\noutput_height <span class=\"token operator\">&#061;<\/span> <span class=\"token punctuation\">(<\/span>input_height <span class=\"token operator\">&#8211;<\/span> kernel_size <span class=\"token operator\">&#043;<\/span> <span class=\"token number\">2<\/span> <span class=\"token operator\">*<\/span> padding<span class=\"token punctuation\">)<\/span> <span class=\"token operator\">\/\/<\/span> stride <span class=\"token operator\">&#043;<\/span> <span class=\"token number\">1<\/span><br \/>\noutput_width <span class=\"token operator\">&#061;<\/span> <span class=\"token punctuation\">(<\/span>input_width <span class=\"token operator\">&#8211;<\/span> kernel_size <span class=\"token operator\">&#043;<\/span> <span class=\"token number\">2<\/span> <span class=\"token operator\">*<\/span> padding<span class=\"token punctuation\">)<\/span> <span class=\"token operator\">\/\/<\/span> stride <span class=\"token operator\">&#043;<\/span> <span class=\"token number\">1<\/span><\/p>\n<p>\u5bf9\u4e8e\u6700\u540e\u4e00\u4e2a\u5377\u79ef\u5757&#xff1a;<\/p>\n<p> \u8f93\u5165&#xff1a;8\u00d78\u00d764<br \/>\n\u5377\u79ef&#xff1a;3\u00d73\u5377\u79ef&#xff0c;\u6b65\u957f1&#xff0c;\u586b\u51451 \u2192 \u8f93\u51fa8\u00d78\u00d7128<br \/>\n\u6c60\u5316&#xff1a;2\u00d72\u6700\u5927\u6c60\u5316&#xff0c;\u6b65\u957f2 \u2192 \u8f93\u51fa4\u00d74\u00d7128\n <\/li>\n<li>\n<p>\u5c55\u5e73\u4e0e\u5168\u8fde\u63a5\u5c42\u8f93\u5165&#xff1a;<\/p>\n<p> <span class=\"token comment\"># \u5c55\u5e73\u64cd\u4f5c<\/span><br \/>\nflattened_size <span class=\"token operator\">&#061;<\/span> <span class=\"token number\">128<\/span> <span class=\"token operator\">*<\/span> <span class=\"token number\">4<\/span> <span class=\"token operator\">*<\/span> <span class=\"token number\">4<\/span> <span class=\"token operator\">&#061;<\/span> <span class=\"token number\">2048<\/span><br \/>\n<span class=\"token comment\"># \u7b2c\u4e00\u4e2a\u5168\u8fde\u63a5\u5c42\u8f93\u5165\u7ef4\u5ea6\u4e3a2048<\/span>\n <\/li>\n<\/ul>\n<h5>2. \u8bad\u7ec3\u7b56\u7565&#xff1a;<\/h5>\n<h6>\u4f18\u5316\u5668\u9009\u62e9<\/h6>\n<ul>\n<li>\n<p>Adam\u4f18\u5316\u5668\u7684\u4f18\u52bf&#xff1a;<\/p>\n<ul>\n<li>\u7ed3\u5408\u4e86\u52a8\u91cf\u6cd5\u548cRMSProp\u7684\u4f18\u70b9<\/li>\n<li>\u81ea\u9002\u5e94\u5b66\u4e60\u7387&#xff0c;\u5bf9\u4e0d\u540c\u53c2\u6570\u4f7f\u7528\u4e0d\u540c\u7684\u5b66\u4e60\u7387<\/li>\n<li>\u5bf9\u8d85\u53c2\u6570\u4e0d\u654f\u611f&#xff0c;\u8bad\u7ec3\u7a33\u5b9a\u6027\u597d<\/li>\n<li>\u9002\u5408\u975e\u51f8\u4f18\u5316\u95ee\u9898\u548c\u5927\u89c4\u6a21\u6570\u636e\u8bad\u7ec3<\/li>\n<\/ul>\n<\/li>\n<li>\n<p>\u5b66\u4e60\u7387\u8bbe\u7f6e&#xff1a;<\/p>\n<ul>\n<li>\u521d\u59cb\u5b66\u4e60\u73871e-3&#xff0c;\u8fd9\u662fAdam\u4f18\u5316\u5668\u7684\u5e38\u7528\u521d\u59cb\u503c<\/li>\n<li>\u8003\u8651\u5230\u6a21\u578b\u89c4\u6a21\u9002\u4e2d&#xff0c;\u672a\u4f7f\u7528\u5b66\u4e60\u7387\u8870\u51cf\u7b56\u7565<\/li>\n<li>\u53ef\u8fdb\u4e00\u6b65\u4f18\u5316&#xff1a;\u4f7f\u7528\u5b66\u4e60\u7387\u8870\u51cf\u6216\u4f59\u5f26\u9000\u706b\u7b56\u7565<\/li>\n<\/ul>\n<\/li>\n<\/ul>\n<h6>\u635f\u5931\u51fd\u6570\u9009\u62e9<\/h6>\n<ul>\n<li>\u4ea4\u53c9\u71b5\u635f\u5931\u7684\u9002\u7528\u6027&#xff1a;\n<ul>\n<li>\u9002\u5408\u591a\u5206\u7c7b\u95ee\u9898&#xff0c;\u76f4\u63a5\u4f18\u5316\u5206\u7c7b\u6982\u7387<\/li>\n<li>\u5bf9\u9519\u8bef\u5206\u7c7b\u7684\u60e9\u7f5a\u66f4\u91cd&#xff0c;\u52a0\u901f\u6a21\u578b\u6536\u655b<\/li>\n<li>PyTorch\u7684nn.CrossEntropyLoss\u7ed3\u5408\u4e86softmax\u548c\u4ea4\u53c9\u71b5&#xff0c;\u4f7f\u7528\u65b9\u4fbf<\/li>\n<\/ul>\n<\/li>\n<\/ul>\n<h6>\u6570\u636e\u589e\u5f3a\u6280\u672f<\/h6>\n<ul>\n<li>\n<p>\u5177\u4f53\u589e\u5f3a\u65b9\u6cd5&#xff1a;<\/p>\n<p> transforms<span class=\"token punctuation\">.<\/span>Compose<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">[<\/span><br \/>\n    transforms<span class=\"token punctuation\">.<\/span>RandomHorizontalFlip<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">,<\/span>  <span class=\"token comment\"># \u968f\u673a\u6c34\u5e73\u7ffb\u8f6c<\/span><br \/>\n    transforms<span class=\"token punctuation\">.<\/span>RandomCrop<span class=\"token punctuation\">(<\/span><span class=\"token number\">32<\/span><span class=\"token punctuation\">,<\/span> padding<span class=\"token operator\">&#061;<\/span><span class=\"token number\">4<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">,<\/span>  <span class=\"token comment\"># \u968f\u673a\u88c1\u526a<\/span><br \/>\n    transforms<span class=\"token punctuation\">.<\/span>ToTensor<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">,<\/span>  <span class=\"token comment\"># \u8f6c\u6362\u4e3a\u5f20\u91cf<\/span><br \/>\n    transforms<span class=\"token punctuation\">.<\/span>Normalize<span class=\"token punctuation\">(<\/span>mean<span class=\"token operator\">&#061;<\/span><span class=\"token punctuation\">[<\/span><span class=\"token number\">0.5<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token number\">0.5<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token number\">0.5<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">,<\/span> std<span class=\"token operator\">&#061;<\/span><span class=\"token punctuation\">[<\/span><span class=\"token number\">0.5<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token number\">0.5<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token number\">0.5<\/span><span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">)<\/span>  <span class=\"token comment\"># \u6807\u51c6\u5316<\/span><br \/>\n<span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">)<\/span>\n <\/li>\n<li>\n<p>\u589e\u5f3a\u6548\u679c&#xff1a;<\/p>\n<ul>\n<li>\u589e\u52a0\u8bad\u7ec3\u6570\u636e\u591a\u6837\u6027&#xff0c;\u51cf\u5c11\u8fc7\u62df\u5408<\/li>\n<li>\u63d0\u9ad8\u6a21\u578b\u7684\u6cdb\u5316\u80fd\u529b\u548c\u9c81\u68d2\u6027<\/li>\n<li>\u6a21\u62df\u771f\u5b9e\u4e16\u754c\u4e2d\u7684\u56fe\u50cf\u53d8\u5316<\/li>\n<\/ul>\n<\/li>\n<\/ul>\n<h6>\u9a8c\u8bc1\u96c6\u76d1\u63a7<\/h6>\n<ul>\n<li>\n<p>\u9a8c\u8bc1\u96c6\u5212\u5206&#xff1a;<\/p>\n<ul>\n<li>\u4ece\u8bad\u7ec3\u96c6\u4e2d\u5212\u520610%\u4f5c\u4e3a\u9a8c\u8bc1\u96c6<\/li>\n<li>\u4f7f\u7528torch.utils.data.random_split\u5b9e\u73b0<\/li>\n<\/ul>\n<\/li>\n<li>\n<p>\u76d1\u63a7\u6307\u6807&#xff1a;<\/p>\n<ul>\n<li>\u8bad\u7ec3\u635f\u5931\u548c\u9a8c\u8bc1\u635f\u5931<\/li>\n<li>\u8bad\u7ec3\u51c6\u786e\u7387\u548c\u9a8c\u8bc1\u51c6\u786e\u7387\n<ul>\n<li>\u8bad\u7ec3\u51c6\u786e\u7387&#xff08;Training Accuracy&#xff09;\n<ul>\n<li>\u5b9a\u4e49&#xff1a;\u6a21\u578b\u5728\u7528\u4e8e\u8bad\u7ec3\u7684\u6570\u636e\u96c6\u4e0a\u6b63\u786e\u5206\u7c7b\u7684\u6837\u672c\u6570\u5360\u603b\u8bad\u7ec3\u6837\u672c\u6570\u7684\u6bd4\u4f8b\u3002<\/li>\n<li>\u6570\u636e\u6765\u6e90&#xff1a;\u6765\u81ea\u8bad\u7ec3\u96c6&#xff08;\u5373\u4ee3\u7801\u4e2dtrainloader\u52a0\u8f7d\u7684\u6570\u636e&#xff09;&#xff0c;\u8fd9\u4e9b\u6570\u636e\u7528\u4e8e\u66f4\u65b0\u6a21\u578b\u7684\u53c2\u6570\u3002<\/li>\n<li>\u4f5c\u7528&#xff1a;\u53cd\u6620\u6a21\u578b\u5bf9\u8bad\u7ec3\u6570\u636e\u7684\u5b66\u4e60\u7a0b\u5ea6\u3002\u8bad\u7ec3\u51c6\u786e\u7387\u9ad8\u8868\u660e\u6a21\u578b\u80fd\u5f88\u597d\u5730\u8bb0\u5fc6\u548c\u62df\u5408\u8bad\u7ec3\u6837\u672c\u3002<\/li>\n<\/ul>\n<\/li>\n<li>\u9a8c\u8bc1\u51c6\u786e\u7387&#xff08;Validation Accuracy&#xff09;\n<ul>\n<li>\u5b9a\u4e49&#xff1a;\u6a21\u578b\u5728\u9a8c\u8bc1\u6570\u636e\u96c6\u4e0a\u6b63\u786e\u5206\u7c7b\u7684\u6837\u672c\u6570\u5360\u603b\u9a8c\u8bc1\u6837\u672c\u6570\u7684\u6bd4\u4f8b\u3002<\/li>\n<li>\u6570\u636e\u6765\u6e90&#xff1a;\u6765\u81ea\u9a8c\u8bc1\u96c6&#xff08;\u5373\u4ee3\u7801\u4e2dvalloader\u52a0\u8f7d\u7684\u6570\u636e&#xff09;&#xff0c;\u8fd9\u4e9b\u6570\u636e\u5728\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u4e0d\u53c2\u4e0e\u53c2\u6570\u66f4\u65b0\u3002<\/li>\n<li>\u4f5c\u7528&#xff1a;\u53cd\u6620\u6a21\u578b\u7684\u6cdb\u5316\u80fd\u529b&#xff0c;\u5373\u6a21\u578b\u5bf9\u672a\u89c1\u8fc7\u7684\u6570\u636e\u7684\u9884\u6d4b\u80fd\u529b\u3002\u9a8c\u8bc1\u51c6\u786e\u7387\u80fd\u5e2e\u52a9\u68c0\u6d4b\u6a21\u578b\u662f\u5426\u8fc7\u62df\u5408&#xff08;\u8bad\u7ec3\u51c6\u786e\u7387\u5f88\u9ad8\u4f46\u9a8c\u8bc1\u51c6\u786e\u7387\u4f4e\u7684\u60c5\u51b5&#xff09;\u3002<\/li>\n<\/ul>\n<\/li>\n<li>\u4e24\u8005\u7684\u5173\u7cfb\u4e0e\u610f\u4e49\n<ul>\n<li>\u5728\u7406\u60f3\u60c5\u51b5\u4e0b&#xff0c;\u8bad\u7ec3\u51c6\u786e\u7387\u548c\u9a8c\u8bc1\u51c6\u786e\u7387\u5e94\u8be5\u90fd\u8f83\u9ad8\u4e14\u5dee\u8ddd\u8f83\u5c0f&#xff0c;\u8fd9\u8868\u660e\u6a21\u578b\u65e2\u5b66\u4e60\u4e86\u8bad\u7ec3\u6570\u636e\u7684\u89c4\u5f8b&#xff0c;\u53c8\u80fd\u5f88\u597d\u5730\u6cdb\u5316\u5230\u65b0\u6570\u636e\u3002<\/li>\n<li>\u5982\u679c\u8bad\u7ec3\u51c6\u786e\u7387\u8fdc\u9ad8\u4e8e\u9a8c\u8bc1\u51c6\u786e\u7387&#xff0c;\u53ef\u80fd\u53d1\u751f\u4e86\u8fc7\u62df\u5408&#xff0c;\u6a21\u578b\u8fc7\u5ea6\u8bb0\u5fc6\u4e86\u8bad\u7ec3\u6570\u636e\u7684\u7ec6\u8282&#xff0c;\u800c\u5ffd\u7565\u4e86\u6570\u636e\u7684\u666e\u904d\u89c4\u5f8b\u3002<\/li>\n<li>\u5982\u679c\u4e24\u8005\u90fd\u5f88\u4f4e&#xff0c;\u53ef\u80fd\u53d1\u751f\u4e86\u6b20\u62df\u5408&#xff0c;\u6a21\u578b\u6ca1\u6709\u5145\u5206\u5b66\u4e60\u5230\u6570\u636e\u7684\u89c4\u5f8b\u3002<\/li>\n<\/ul>\n<\/li>\n<\/ul>\n<\/li>\n<li>\u53ef\u8fdb\u4e00\u6b65\u6dfb\u52a0\u5176\u4ed6\u6307\u6807&#xff0c;\u5982\u7cbe\u786e\u7387\u3001\u53ec\u56de\u7387\u7b49<\/li>\n<\/ul>\n<\/li>\n<li>\n<p>\u65e9\u505c\u7b56\u7565&#xff1a;<\/p>\n<ul>\n<li>\u53ef\u5b9e\u73b0\u65e9\u505c\u7b56\u7565&#xff0c;\u5f53\u9a8c\u8bc1\u635f\u5931\u8fde\u7eed\u591a\u8f6e\u4e0d\u4e0b\u964d\u65f6\u505c\u6b62\u8bad\u7ec3<\/li>\n<li>\u4fdd\u5b58\u9a8c\u8bc1\u635f\u5931\u6700\u4f4e\u7684\u6a21\u578b<\/li>\n<\/ul>\n<\/li>\n<\/ul>\n<h5>3. \u53ef\u89c6\u5316\u8bbe\u8ba1&#xff1a;<\/h5>\n<h6>\u8bad\u7ec3\u635f\u5931\u66f2\u7ebf<\/h6>\n<ul>\n<li>\n<p>\u5b9e\u73b0\u65b9\u6cd5&#xff1a;<\/p>\n<ul>\n<li>\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u8bb0\u5f55\u6bcf\u8f6e\u7684\u8bad\u7ec3\u635f\u5931\u548c\u9a8c\u8bc1\u635f\u5931<\/li>\n<li>\u4f7f\u7528matplotlib\u7ed8\u5236\u635f\u5931\u66f2\u7ebf<\/li>\n<\/ul>\n<\/li>\n<li>\n<p>\u6280\u672f\u7ec6\u8282&#xff1a;<\/p>\n<p> plt<span class=\"token punctuation\">.<\/span>figure<span class=\"token punctuation\">(<\/span>figsize<span class=\"token operator\">&#061;<\/span><span class=\"token punctuation\">(<\/span><span class=\"token number\">10<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token number\">5<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">)<\/span><br \/>\nplt<span class=\"token punctuation\">.<\/span>plot<span class=\"token punctuation\">(<\/span><span class=\"token builtin\">range<\/span><span class=\"token punctuation\">(<\/span>epochs<span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">,<\/span> train_losses<span class=\"token punctuation\">,<\/span> label<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#039;Training Loss&#039;<\/span><span class=\"token punctuation\">)<\/span><br \/>\nplt<span class=\"token punctuation\">.<\/span>plot<span class=\"token punctuation\">(<\/span><span class=\"token builtin\">range<\/span><span class=\"token punctuation\">(<\/span>epochs<span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">,<\/span> val_losses<span class=\"token punctuation\">,<\/span> label<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#039;Validation Loss&#039;<\/span><span class=\"token punctuation\">)<\/span><br \/>\nplt<span class=\"token punctuation\">.<\/span>xlabel<span class=\"token punctuation\">(<\/span><span class=\"token string\">&#039;Epochs&#039;<\/span><span class=\"token punctuation\">)<\/span><br \/>\nplt<span class=\"token punctuation\">.<\/span>ylabel<span class=\"token punctuation\">(<\/span><span class=\"token string\">&#039;Loss&#039;<\/span><span class=\"token punctuation\">)<\/span><br \/>\nplt<span class=\"token punctuation\">.<\/span>legend<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><br \/>\nplt<span class=\"token punctuation\">.<\/span>grid<span class=\"token punctuation\">(<\/span><span class=\"token boolean\">True<\/span><span class=\"token punctuation\">)<\/span><br \/>\nplt<span class=\"token punctuation\">.<\/span>savefig<span class=\"token punctuation\">(<\/span><span class=\"token string\">&#039;training_loss_curve.png&#039;<\/span><span class=\"token punctuation\">)<\/span>\n <\/li>\n<li>\n<p>\u7ed3\u679c\u89e3\u8bfb&#xff1a;<\/p>\n<ul>\n<li>\u7406\u60f3\u66f2\u7ebf&#xff1a;\u8bad\u7ec3\u635f\u5931\u548c\u9a8c\u8bc1\u635f\u5931\u90fd\u9010\u6e10\u4e0b\u964d&#xff0c;\u6700\u7ec8\u8d8b\u4e8e\u7a33\u5b9a<\/li>\n<li>\u8fc7\u62df\u5408&#xff1a;\u8bad\u7ec3\u635f\u5931\u6301\u7eed\u4e0b\u964d&#xff0c;\u9a8c\u8bc1\u635f\u5931\u5148\u964d\u540e\u5347<\/li>\n<li>\u6b20\u62df\u5408&#xff1a;\u8bad\u7ec3\u635f\u5931\u548c\u9a8c\u8bc1\u635f\u5931\u90fd\u5f88\u9ad8<\/li>\n<\/ul>\n<\/li>\n<\/ul>\n<h6>\u6df7\u6dc6\u77e9\u9635<\/h6>\n<ul>\n<li>\n<p>\u5b9e\u73b0\u65b9\u6cd5&#xff1a;<\/p>\n<ul>\n<li>\u4f7f\u7528sklearn\u7684confusion_matrix\u51fd\u6570\u8ba1\u7b97<\/li>\n<li>\u4f7f\u7528seaborn\u7684heatmap\u51fd\u6570\u53ef\u89c6\u5316<\/li>\n<\/ul>\n<\/li>\n<li>\n<p>\u6280\u672f\u7ec6\u8282&#xff1a;<\/p>\n<p> <span class=\"token keyword\">from<\/span> sklearn<span class=\"token punctuation\">.<\/span>metrics <span class=\"token keyword\">import<\/span> confusion_matrix<br \/>\n<span class=\"token keyword\">import<\/span> seaborn <span class=\"token keyword\">as<\/span> sns<\/p>\n<p>cm <span class=\"token operator\">&#061;<\/span> confusion_matrix<span class=\"token punctuation\">(<\/span>true_labels<span class=\"token punctuation\">,<\/span> predicted_labels<span class=\"token punctuation\">)<\/span><br \/>\nplt<span class=\"token punctuation\">.<\/span>figure<span class=\"token punctuation\">(<\/span>figsize<span class=\"token operator\">&#061;<\/span><span class=\"token punctuation\">(<\/span><span class=\"token number\">12<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token number\">10<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">)<\/span><br \/>\nsns<span class=\"token punctuation\">.<\/span>heatmap<span class=\"token punctuation\">(<\/span>cm<span class=\"token punctuation\">,<\/span> annot<span class=\"token operator\">&#061;<\/span><span class=\"token boolean\">True<\/span><span class=\"token punctuation\">,<\/span> fmt<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#039;d&#039;<\/span><span class=\"token punctuation\">,<\/span> cmap<span class=\"token operator\">&#061;<\/span><span class=\"token string\">&#039;Blues&#039;<\/span><span class=\"token punctuation\">)<\/span><br \/>\nplt<span class=\"token punctuation\">.<\/span>xlabel<span class=\"token punctuation\">(<\/span><span class=\"token string\">&#039;Predicted Label&#039;<\/span><span class=\"token punctuation\">)<\/span><br \/>\nplt<span class=\"token punctuation\">.<\/span>ylabel<span class=\"token punctuation\">(<\/span><span class=\"token string\">&#039;True Label&#039;<\/span><span class=\"token punctuation\">)<\/span><br \/>\nplt<span class=\"token punctuation\">.<\/span>savefig<span class=\"token punctuation\">(<\/span><span class=\"token string\">&#039;confusion_matrix.png&#039;<\/span><span class=\"token punctuation\">)<\/span>\n <\/li>\n<li>\n<p>\u7ed3\u679c\u89e3\u8bfb&#xff1a;<\/p>\n<ul>\n<li>\u5bf9\u89d2\u7ebf\u5143\u7d20\u8868\u793a\u6b63\u786e\u5206\u7c7b\u7684\u6837\u672c\u6570<\/li>\n<li>\u975e\u5bf9\u89d2\u7ebf\u5143\u7d20\u8868\u793a\u9519\u8bef\u5206\u7c7b\u7684\u6837\u672c\u6570<\/li>\n<li>\u53ef\u76f4\u89c2\u770b\u51fa\u54ea\u4e9b\u7c7b\u522b\u5bb9\u6613\u6df7\u6dc6<\/li>\n<\/ul>\n<\/li>\n<\/ul>\n<h6>\u6837\u672c\u5c55\u793a<\/h6>\n<ul>\n<li>\n<p>\u9519\u8bef\u5206\u7c7b\u6837\u672c&#xff1a;<\/p>\n<ul>\n<li>\u5c55\u793a\u6a21\u578b\u5206\u7c7b\u9519\u8bef\u7684\u6837\u672c&#xff0c;\u5e2e\u52a9\u5206\u6790\u6a21\u578b\u7684\u5f31\u70b9<\/li>\n<li>\u663e\u793a\u771f\u5b9e\u6807\u7b7e\u548c\u9884\u6d4b\u6807\u7b7e<\/li>\n<\/ul>\n<\/li>\n<li>\n<p>\u6b63\u786e\u5206\u7c7b\u6837\u672c&#xff1a;<\/p>\n<ul>\n<li>\u5c55\u793a\u6a21\u578b\u5206\u7c7b\u6b63\u786e\u7684\u6837\u672c&#xff0c;\u7279\u522b\u662f\u7f6e\u4fe1\u5ea6\u8f83\u4f4e\u7684\u6837\u672c<\/li>\n<li>\u5e2e\u52a9\u4e86\u89e3\u6a21\u578b\u7684\u4f18\u52bf<\/li>\n<\/ul>\n<\/li>\n<li>\n<p>\u6280\u672f\u7ec6\u8282&#xff1a;<\/p>\n<p> <span class=\"token comment\"># \u663e\u793a\u524d25\u4e2a\u9519\u8bef\u5206\u7c7b\u6837\u672c<\/span><br \/>\nplt<span class=\"token punctuation\">.<\/span>figure<span class=\"token punctuation\">(<\/span>figsize<span class=\"token operator\">&#061;<\/span><span class=\"token punctuation\">(<\/span><span class=\"token number\">10<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token number\">10<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">)<\/span><br \/>\n<span class=\"token keyword\">for<\/span> i <span class=\"token keyword\">in<\/span> <span class=\"token builtin\">range<\/span><span class=\"token punctuation\">(<\/span><span class=\"token number\">25<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">:<\/span><br \/>\n    plt<span class=\"token punctuation\">.<\/span>subplot<span class=\"token punctuation\">(<\/span><span class=\"token number\">5<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token number\">5<\/span><span class=\"token punctuation\">,<\/span> i<span class=\"token operator\">&#043;<\/span><span class=\"token number\">1<\/span><span class=\"token punctuation\">)<\/span><br \/>\n    plt<span class=\"token punctuation\">.<\/span>imshow<span class=\"token punctuation\">(<\/span>misclassified_images<span class=\"token punctuation\">[<\/span>i<span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">.<\/span>permute<span class=\"token punctuation\">(<\/span><span class=\"token number\">1<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token number\">2<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token number\">0<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">)<\/span><br \/>\n    plt<span class=\"token punctuation\">.<\/span>title<span class=\"token punctuation\">(<\/span><span class=\"token string-interpolation\"><span class=\"token string\">f&#039;True: <\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span>true_label<span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\">, Pred: <\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span>pred_label<span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\">&#039;<\/span><\/span><span class=\"token punctuation\">)<\/span><br \/>\n    plt<span class=\"token punctuation\">.<\/span>axis<span class=\"token punctuation\">(<\/span><span class=\"token string\">&#039;off&#039;<\/span><span class=\"token punctuation\">)<\/span><br \/>\nplt<span class=\"token punctuation\">.<\/span>tight_layout<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><br \/>\nplt<span class=\"token punctuation\">.<\/span>savefig<span class=\"token punctuation\">(<\/span><span class=\"token string\">&#039;misclassified_samples.png&#039;<\/span><span class=\"token punctuation\">)<\/span>\n <\/li>\n<\/ul>\n<h6>\u9884\u6d4b\u6982\u7387\u5206\u5e03<\/h6>\n<ul>\n<li>\n<p>\u5b9e\u73b0\u65b9\u6cd5&#xff1a;<\/p>\n<ul>\n<li>\u4f7f\u7528\u6a21\u578b\u7684softmax\u8f93\u51fa\u83b7\u53d6\u9884\u6d4b\u6982\u7387<\/li>\n<li>\u7ed8\u5236\u6982\u7387\u5206\u5e03\u76f4\u65b9\u56fe\u6216\u6761\u5f62\u56fe<\/li>\n<\/ul>\n<\/li>\n<li>\n<p>\u6280\u672f\u7ec6\u8282&#xff1a;<\/p>\n<p> <span class=\"token comment\"># \u83b7\u53d6\u9884\u6d4b\u6982\u7387<\/span><br \/>\nprobabilities <span class=\"token operator\">&#061;<\/span> F<span class=\"token punctuation\">.<\/span>softmax<span class=\"token punctuation\">(<\/span>model<span class=\"token punctuation\">(<\/span>inputs<span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">,<\/span> dim<span class=\"token operator\">&#061;<\/span><span class=\"token number\">1<\/span><span class=\"token punctuation\">)<\/span><br \/>\n<span class=\"token comment\"># \u7ed8\u5236\u524d5\u4e2a\u6837\u672c\u7684\u6982\u7387\u5206\u5e03<\/span><br \/>\nplt<span class=\"token punctuation\">.<\/span>figure<span class=\"token punctuation\">(<\/span>figsize<span class=\"token operator\">&#061;<\/span><span class=\"token punctuation\">(<\/span><span class=\"token number\">15<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token number\">5<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">)<\/span><br \/>\n<span class=\"token keyword\">for<\/span> i <span class=\"token keyword\">in<\/span> <span class=\"token builtin\">range<\/span><span class=\"token punctuation\">(<\/span><span class=\"token number\">5<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">:<\/span><br \/>\n    plt<span class=\"token punctuation\">.<\/span>subplot<span class=\"token punctuation\">(<\/span><span class=\"token number\">1<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token number\">5<\/span><span class=\"token punctuation\">,<\/span> i<span class=\"token operator\">&#043;<\/span><span class=\"token number\">1<\/span><span class=\"token punctuation\">)<\/span><br \/>\n    plt<span class=\"token punctuation\">.<\/span>bar<span class=\"token punctuation\">(<\/span><span class=\"token builtin\">range<\/span><span class=\"token punctuation\">(<\/span><span class=\"token number\">10<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">,<\/span> probabilities<span class=\"token punctuation\">[<\/span>i<span class=\"token punctuation\">]<\/span><span class=\"token punctuation\">.<\/span>detach<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">.<\/span>numpy<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">)<\/span><br \/>\n    plt<span class=\"token punctuation\">.<\/span>title<span class=\"token punctuation\">(<\/span><span class=\"token string-interpolation\"><span class=\"token string\">f&#039;Sample <\/span><span class=\"token interpolation\"><span class=\"token punctuation\">{<\/span>i<span class=\"token operator\">&#043;<\/span><span class=\"token number\">1<\/span><span class=\"token punctuation\">}<\/span><\/span><span class=\"token string\">&#039;<\/span><\/span><span class=\"token punctuation\">)<\/span><br \/>\n    plt<span class=\"token punctuation\">.<\/span>xlabel<span class=\"token punctuation\">(<\/span><span class=\"token string\">&#039;Class&#039;<\/span><span class=\"token punctuation\">)<\/span><br \/>\n    plt<span class=\"token punctuation\">.<\/span>ylabel<span class=\"token punctuation\">(<\/span><span class=\"token string\">&#039;Probability&#039;<\/span><span class=\"token punctuation\">)<\/span><br \/>\nplt<span class=\"token punctuation\">.<\/span>tight_layout<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><br \/>\nplt<span class=\"token punctuation\">.<\/span>savefig<span class=\"token punctuation\">(<\/span><span class=\"token string\">&#039;prediction_probabilities.png&#039;<\/span><span class=\"token punctuation\">)<\/span>\n <\/li>\n<li>\n<p>\u7ed3\u679c\u89e3\u8bfb&#xff1a;<\/p>\n<ul>\n<li>\u89c2\u5bdf\u6a21\u578b\u5bf9\u6b63\u786e\u7c7b\u522b\u7684\u7f6e\u4fe1\u5ea6<\/li>\n<li>\u5206\u6790\u6a21\u578b\u662f\u5426\u5bf9\u9519\u8bef\u7c7b\u522b\u6709\u8f83\u9ad8\u7684\u7f6e\u4fe1\u5ea6<\/li>\n<li>\u4e86\u89e3\u6a21\u578b\u7684\u51b3\u7b56\u8fb9\u754c<\/li>\n<\/ul>\n<\/li>\n<\/ul>\n<h4>4.3 \u7ed3\u5408\u77e5\u8bc6\u70b9\u4e0e\u516c\u5f0f\u7684\u4ee3\u7801\u5206\u6790<\/h4>\n<h5>4.3.1 \u5377\u79ef\u64cd\u4f5c&#xff1a;\u56fe\u50cf\u7279\u5f81\u63d0\u53d6\u7684\u6570\u5b66\u539f\u7406<\/h5>\n<p>nn<span class=\"token punctuation\">.<\/span>Conv2d<span class=\"token punctuation\">(<\/span><span class=\"token number\">3<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token number\">32<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token number\">3<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token number\">1<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token number\">1<\/span><span class=\"token punctuation\">)<\/span>  <span class=\"token comment\"># 3\u901a\u9053\u8f93\u5165&#xff0c;32\u901a\u9053\u8f93\u51fa&#xff0c;3\u00d73\u5377\u79ef\u6838&#xff0c;\u6b65\u957f1&#xff0c;\u586b\u51451<\/span><\/p>\n<h6>\u6570\u5b66\u539f\u7406\u4e0e\u63a8\u5bfc&#xff1a;<\/h6>\n<p>\u5377\u79ef\u50cf\u64cd\u4f5c\u7684\u672c\u8d28\u662f\u5728\u8f93\u5165\u7279\u5f81\u56fe\u4e0a\u6ed1\u52a8\u5377\u79ef\u6838&#xff0c;\u8ba1\u7b97\u5c40\u90e8\u533a\u57df\u7684\u52a0\u6743\u548c\u3002\u5bf9\u4e8e\u591a\u901a\u9053\u8f93\u5165&#xff0c;\u6bcf\u4e2a\u901a\u9053\u90fd\u6709\u72ec\u7acb\u7684\u5377\u79ef\u6838&#xff0c;\u8f93\u51fa\u662f\u6240\u6709\u901a\u9053\u7ed3\u679c\u7684\u603b\u548c\u3002<\/p>\n<p>\u591a\u901a\u9053\u5377\u79ef\u516c\u5f0f&#xff1a; <span class=\"katex--display\"><span class=\"katex-display\"><span class=\"katex\"><span class=\"katex-mathml\"><\/p>\n<p>         (<\/p>\n<p>         f<\/p>\n<p>         \u2217<\/p>\n<p>         g<\/p>\n<p>         )<\/p>\n<p>         (<\/p>\n<p>         i<\/p>\n<p>         ,<\/p>\n<p>         j<\/p>\n<p>         ,<\/p>\n<p>         k<\/p>\n<p>         )<\/p>\n<p>         &#061;<\/p>\n<p>          \u2211<\/p>\n<p>           c<\/p>\n<p>           &#061;<\/p>\n<p>           1<\/p>\n<p>          C<\/p>\n<p>          \u2211<\/p>\n<p>           m<\/p>\n<p>           &#061;<\/p>\n<p>           \u2212<\/p>\n<p>           1<\/p>\n<p>          1<\/p>\n<p>          \u2211<\/p>\n<p>           n<\/p>\n<p>           &#061;<\/p>\n<p>           \u2212<\/p>\n<p>           1<\/p>\n<p>          1<\/p>\n<p>         f<\/p>\n<p>         (<\/p>\n<p>         c<\/p>\n<p>         ,<\/p>\n<p>         m<\/p>\n<p>         ,<\/p>\n<p>         n<\/p>\n<p>         ,<\/p>\n<p>         k<\/p>\n<p>         )<\/p>\n<p>         g<\/p>\n<p>         (<\/p>\n<p>         i<\/p>\n<p>         &#043;<\/p>\n<p>         m<\/p>\n<p>         ,<\/p>\n<p>         j<\/p>\n<p>         &#043;<\/p>\n<p>         n<\/p>\n<p>         ,<\/p>\n<p>         c<\/p>\n<p>         )<\/p>\n<p>         (f * g)(i, j, k) &#061; \\\\sum_{c&#061;1}^{C} \\\\sum_{m&#061;-1}^{1} \\\\sum_{n&#061;-1}^{1} f(c, m, n, k) g(i&#043;m, j&#043;n, c) <\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 1em;vertical-align: -0.25em\"><\/span><span class=\"mopen\">(<\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.1076em\">f<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u2217<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 1em;vertical-align: -0.25em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0359em\">g<\/span><span class=\"mclose\">)<\/span><span class=\"mopen\">(<\/span><span class=\"mord mathnormal\">i<\/span><span class=\"mpunct\">,<\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0572em\">j<\/span><span class=\"mpunct\">,<\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0315em\">k<\/span><span class=\"mclose\">)<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 3.1538em;vertical-align: -1.3254em\"><\/span><span class=\"mop op-limits\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.8283em\"><span class=\"\" style=\"top: -1.8829em;margin-left: 0em\"><span class=\"pstrut\" style=\"height: 3.05em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mathnormal mtight\">c<\/span><span class=\"mrel mtight\">&#061;<\/span><span class=\"mord mtight\">1<\/span><\/span><\/span><\/span><span class=\"\" style=\"top: -3.05em\"><span class=\"pstrut\" style=\"height: 3.05em\"><\/span><span class=\"\"><span class=\"mop op-symbol large-op\">\u2211<\/span><\/span><\/span><span class=\"\" style=\"top: -4.3em;margin-left: 0em\"><span class=\"pstrut\" style=\"height: 3.05em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mathnormal mtight\" style=\"margin-right: 0.0715em\">C<\/span><\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.2671em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mop op-limits\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.8011em\"><span class=\"\" style=\"top: -1.8829em;margin-left: 0em\"><span class=\"pstrut\" style=\"height: 3.05em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mathnormal mtight\">m<\/span><span class=\"mrel mtight\">&#061;<\/span><span class=\"mord mtight\">\u2212<\/span><span class=\"mord mtight\">1<\/span><\/span><\/span><\/span><span class=\"\" style=\"top: -3.05em\"><span class=\"pstrut\" style=\"height: 3.05em\"><\/span><span class=\"\"><span class=\"mop op-symbol large-op\">\u2211<\/span><\/span><\/span><span class=\"\" style=\"top: -4.3em;margin-left: 0em\"><span class=\"pstrut\" style=\"height: 3.05em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mtight\">1<\/span><\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.3254em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mop op-limits\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.8011em\"><span class=\"\" style=\"top: -1.8829em;margin-left: 0em\"><span class=\"pstrut\" style=\"height: 3.05em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mathnormal mtight\">n<\/span><span class=\"mrel mtight\">&#061;<\/span><span class=\"mord mtight\">\u2212<\/span><span class=\"mord mtight\">1<\/span><\/span><\/span><\/span><span class=\"\" style=\"top: -3.05em\"><span class=\"pstrut\" style=\"height: 3.05em\"><\/span><span class=\"\"><span class=\"mop op-symbol large-op\">\u2211<\/span><\/span><\/span><span class=\"\" style=\"top: -4.3em;margin-left: 0em\"><span class=\"pstrut\" style=\"height: 3.05em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mtight\">1<\/span><\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.3254em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.1076em\">f<\/span><span class=\"mopen\">(<\/span><span class=\"mord mathnormal\">c<\/span><span class=\"mpunct\">,<\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord mathnormal\">m<\/span><span class=\"mpunct\">,<\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord mathnormal\">n<\/span><span class=\"mpunct\">,<\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0315em\">k<\/span><span class=\"mclose\">)<\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0359em\">g<\/span><span class=\"mopen\">(<\/span><span class=\"mord mathnormal\">i<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">&#043;<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.854em;vertical-align: -0.1944em\"><\/span><span class=\"mord mathnormal\">m<\/span><span class=\"mpunct\">,<\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0572em\">j<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">&#043;<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 1em;vertical-align: -0.25em\"><\/span><span class=\"mord mathnormal\">n<\/span><span class=\"mpunct\">,<\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord mathnormal\">c<\/span><span class=\"mclose\">)<\/span><\/span><\/span><\/span><\/span><\/span><\/p>\n<p>\u5176\u4e2d&#xff1a;<\/p>\n<ul>\n<li><span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">\n<p>         C<\/p>\n<p>        C<\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.6833em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0715em\">C<\/span><\/span><\/span><\/span><\/span>&#xff1a;\u8f93\u5165\u901a\u9053\u6570&#xff08;\u6b64\u5904\u4e3a3&#xff09;<\/li>\n<li><span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">\n<p>         k<\/p>\n<p>        k<\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.6944em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0315em\">k<\/span><\/span><\/span><\/span><\/span>&#xff1a;\u8f93\u51fa\u901a\u9053\u6570&#xff08;\u6b64\u5904\u4e3a32&#xff09;<\/li>\n<li><span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">\n<p>         f<\/p>\n<p>         (<\/p>\n<p>         c<\/p>\n<p>         ,<\/p>\n<p>         m<\/p>\n<p>         ,<\/p>\n<p>         n<\/p>\n<p>         ,<\/p>\n<p>         k<\/p>\n<p>         )<\/p>\n<p>        f(c, m, n, k)<\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 1em;vertical-align: -0.25em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.1076em\">f<\/span><span class=\"mopen\">(<\/span><span class=\"mord mathnormal\">c<\/span><span class=\"mpunct\">,<\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord mathnormal\">m<\/span><span class=\"mpunct\">,<\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord mathnormal\">n<\/span><span class=\"mpunct\">,<\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0315em\">k<\/span><span class=\"mclose\">)<\/span><\/span><\/span><\/span><\/span>&#xff1a;\u7b2c<span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\"><\/p>\n<p>         k<\/p>\n<p>        k<\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.6944em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0315em\">k<\/span><\/span><\/span><\/span><\/span>\u4e2a\u8f93\u51fa\u901a\u9053\u5bf9\u5e94\u7684\u7b2c<span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\"><\/p>\n<p>         c<\/p>\n<p>        c<\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.4306em\"><\/span><span class=\"mord mathnormal\">c<\/span><\/span><\/span><\/span><\/span>\u4e2a\u8f93\u5165\u901a\u9053\u7684\u5377\u79ef\u6838\u6743\u91cd<\/li>\n<li><span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">\n<p>         g<\/p>\n<p>         (<\/p>\n<p>         i<\/p>\n<p>         &#043;<\/p>\n<p>         m<\/p>\n<p>         ,<\/p>\n<p>         j<\/p>\n<p>         &#043;<\/p>\n<p>         n<\/p>\n<p>         ,<\/p>\n<p>         c<\/p>\n<p>         )<\/p>\n<p>        g(i&#043;m, j&#043;n, c)<\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 1em;vertical-align: -0.25em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0359em\">g<\/span><span class=\"mopen\">(<\/span><span class=\"mord mathnormal\">i<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">&#043;<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.854em;vertical-align: -0.1944em\"><\/span><span class=\"mord mathnormal\">m<\/span><span class=\"mpunct\">,<\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0572em\">j<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">&#043;<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 1em;vertical-align: -0.25em\"><\/span><span class=\"mord mathnormal\">n<\/span><span class=\"mpunct\">,<\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord mathnormal\">c<\/span><span class=\"mclose\">)<\/span><\/span><\/span><\/span><\/span>&#xff1a;\u8f93\u5165\u7279\u5f81\u56fe\u5728\u4f4d\u7f6e<span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\"><\/p>\n<p>         (<\/p>\n<p>         i<\/p>\n<p>         &#043;<\/p>\n<p>         m<\/p>\n<p>         ,<\/p>\n<p>         j<\/p>\n<p>         &#043;<\/p>\n<p>         n<\/p>\n<p>         )<\/p>\n<p>        (i&#043;m, j&#043;n)<\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 1em;vertical-align: -0.25em\"><\/span><span class=\"mopen\">(<\/span><span class=\"mord mathnormal\">i<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">&#043;<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.854em;vertical-align: -0.1944em\"><\/span><span class=\"mord mathnormal\">m<\/span><span class=\"mpunct\">,<\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0572em\">j<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">&#043;<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 1em;vertical-align: -0.25em\"><\/span><span class=\"mord mathnormal\">n<\/span><span class=\"mclose\">)<\/span><\/span><\/span><\/span><\/span>\u5904\u7684\u7b2c<span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\"><\/p>\n<p>         c<\/p>\n<p>        c<\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.4306em\"><\/span><span class=\"mord mathnormal\">c<\/span><\/span><\/span><\/span><\/span>\u4e2a\u901a\u9053\u503c<\/li>\n<\/ul>\n<p>\u8f93\u51fa\u7279\u5f81\u56fe\u5c3a\u5bf8\u8ba1\u7b97&#xff1a; <span class=\"katex--display\"><span class=\"katex-display\"><span class=\"katex\"><span class=\"katex-mathml\"><\/p>\n<p>         O<\/p>\n<p>         u<\/p>\n<p>         t<\/p>\n<p>         p<\/p>\n<p>         u<\/p>\n<p>         t<\/p>\n<p>         _<\/p>\n<p>         s<\/p>\n<p>         i<\/p>\n<p>         z<\/p>\n<p>         e<\/p>\n<p>         &#061;<\/p>\n<p>           I<\/p>\n<p>           n<\/p>\n<p>           p<\/p>\n<p>           u<\/p>\n<p>           t<\/p>\n<p>           _<\/p>\n<p>           s<\/p>\n<p>           i<\/p>\n<p>           z<\/p>\n<p>           e<\/p>\n<p>           &#043;<\/p>\n<p>           2<\/p>\n<p>           \u2217<\/p>\n<p>           P<\/p>\n<p>           a<\/p>\n<p>           d<\/p>\n<p>           d<\/p>\n<p>           i<\/p>\n<p>           n<\/p>\n<p>           g<\/p>\n<p>           \u2212<\/p>\n<p>           K<\/p>\n<p>           e<\/p>\n<p>           r<\/p>\n<p>           n<\/p>\n<p>           e<\/p>\n<p>           l<\/p>\n<p>           _<\/p>\n<p>           s<\/p>\n<p>           i<\/p>\n<p>           z<\/p>\n<p>           e<\/p>\n<p>           S<\/p>\n<p>           t<\/p>\n<p>           r<\/p>\n<p>           i<\/p>\n<p>           d<\/p>\n<p>           e<\/p>\n<p>         &#043;<\/p>\n<p>         1<\/p>\n<p>         Output\\\\_size &#061; \\\\frac{Input\\\\_size &#043; 2*Padding &#8211; Kernel\\\\_size}{Stride} &#043; 1 <\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.9933em;vertical-align: -0.31em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0278em\">O<\/span><span class=\"mord mathnormal\">u<\/span><span class=\"mord mathnormal\">tp<\/span><span class=\"mord mathnormal\">u<\/span><span class=\"mord mathnormal\">t<\/span><span class=\"mord\" style=\"margin-right: 0.0278em\">_<\/span><span class=\"mord mathnormal\">s<\/span><span class=\"mord mathnormal\">i<\/span><span class=\"mord mathnormal\">ze<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 2.0804em;vertical-align: -0.686em\"><\/span><span class=\"mord\"><span class=\"mopen nulldelimiter\"><\/span><span class=\"mfrac\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.3944em\"><span class=\"\" style=\"top: -2.314em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\">St<\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0278em\">r<\/span><span class=\"mord mathnormal\">i<\/span><span class=\"mord mathnormal\">d<\/span><span class=\"mord mathnormal\">e<\/span><\/span><\/span><span class=\"\" style=\"top: -3.23em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"frac-line\" style=\"border-bottom-width: 0.04em\"><\/span><\/span><span class=\"\" style=\"top: -3.7em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.0785em\">I<\/span><span class=\"mord mathnormal\">n<\/span><span class=\"mord mathnormal\">p<\/span><span class=\"mord mathnormal\">u<\/span><span class=\"mord mathnormal\">t<\/span><span class=\"mord\" style=\"margin-right: 0.0278em\">_<\/span><span class=\"mord mathnormal\">s<\/span><span class=\"mord mathnormal\">i<\/span><span class=\"mord mathnormal\">ze<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">&#043;<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mord\">2<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u2217<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.1389em\">P<\/span><span class=\"mord mathnormal\">a<\/span><span class=\"mord mathnormal\">dd<\/span><span class=\"mord mathnormal\">in<\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0359em\">g<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u2212<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0278em\">Ker<\/span><span class=\"mord mathnormal\">n<\/span><span class=\"mord mathnormal\">e<\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0197em\">l<\/span><span class=\"mord\" style=\"margin-right: 0.0278em\">_<\/span><span class=\"mord mathnormal\">s<\/span><span class=\"mord mathnormal\">i<\/span><span class=\"mord mathnormal\">ze<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.686em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><span class=\"mclose nulldelimiter\"><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">&#043;<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.6444em\"><\/span><span class=\"mord\">1<\/span><\/span><\/span><\/span><\/span><\/span><\/p>\n<p>\u5bf9\u4e8e\u672c\u9879\u76ee&#xff1a; <span class=\"katex--display\"><span class=\"katex-display\"><span class=\"katex\"><span class=\"katex-mathml\"><\/p>\n<p>         O<\/p>\n<p>         u<\/p>\n<p>         t<\/p>\n<p>         p<\/p>\n<p>         u<\/p>\n<p>         t<\/p>\n<p>         _<\/p>\n<p>         s<\/p>\n<p>         i<\/p>\n<p>         z<\/p>\n<p>         e<\/p>\n<p>         &#061;<\/p>\n<p>           32<\/p>\n<p>           &#043;<\/p>\n<p>           2<\/p>\n<p>           \u2217<\/p>\n<p>           1<\/p>\n<p>           \u2212<\/p>\n<p>           3<\/p>\n<p>          1<\/p>\n<p>         &#043;<\/p>\n<p>         1<\/p>\n<p>         &#061;<\/p>\n<p>         32<\/p>\n<p>         Output\\\\_size &#061; \\\\frac{32 &#043; 2*1 &#8211; 3}{1} &#043; 1 &#061; 32 <\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.9933em;vertical-align: -0.31em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0278em\">O<\/span><span class=\"mord mathnormal\">u<\/span><span class=\"mord mathnormal\">tp<\/span><span class=\"mord mathnormal\">u<\/span><span class=\"mord mathnormal\">t<\/span><span class=\"mord\" style=\"margin-right: 0.0278em\">_<\/span><span class=\"mord mathnormal\">s<\/span><span class=\"mord mathnormal\">i<\/span><span class=\"mord mathnormal\">ze<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 2.0074em;vertical-align: -0.686em\"><\/span><span class=\"mord\"><span class=\"mopen nulldelimiter\"><\/span><span class=\"mfrac\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.3214em\"><span class=\"\" style=\"top: -2.314em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"mord\"><span class=\"mord\">1<\/span><\/span><\/span><span class=\"\" style=\"top: -3.23em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"frac-line\" style=\"border-bottom-width: 0.04em\"><\/span><\/span><span class=\"\" style=\"top: -3.677em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"mord\"><span class=\"mord\">32<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">&#043;<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mord\">2<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u2217<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mord\">1<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u2212<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mord\">3<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.686em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><span class=\"mclose nulldelimiter\"><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">&#043;<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.6444em\"><\/span><span class=\"mord\">1<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.6444em\"><\/span><span class=\"mord\">32<\/span><\/span><\/span><\/span><\/span><\/span>&#034; \u4fdd\u6301\u4e86\u8f93\u5165\u8f93\u51fa\u5c3a\u5bf8\u4e0d\u53d8&#xff08;same\u5377\u79ef&#xff09;\u3002<\/p>\n<h6>\u4ee3\u7801\u8bbe\u8ba1\u601d\u8def&#xff1a;<\/h6>\n<ul>\n<li>3&#215;3\u5377\u79ef\u6838&#xff1a;\u5e73\u8861\u4e86\u8ba1\u7b97\u6548\u7387\u548c\u7279\u5f81\u63d0\u53d6\u80fd\u529b&#xff0c;\u662fCNN\u4e2d\u6700\u5e38\u7528\u7684\u5377\u79ef\u6838\u5c3a\u5bf8<\/li>\n<li>32\u4e2a\u8f93\u51fa\u901a\u9053&#xff1a;\u7b2c\u4e00\u5c42\u5377\u79ef\u4f7f\u7528\u8f83\u5c11\u7684\u901a\u9053\u6570&#xff0c;\u907f\u514d\u53c2\u6570\u7206\u70b8<\/li>\n<li>\u6b65\u957f1&#xff1a;\u4fdd\u7559\u66f4\u591a\u7a7a\u95f4\u4fe1\u606f<\/li>\n<li>\u586b\u51451&#xff1a;\u4fdd\u6301\u8f93\u51fa\u5c3a\u5bf8\u4e0e\u8f93\u5165\u76f8\u540c&#xff0c;\u907f\u514d\u8fb9\u7f18\u4fe1\u606f\u4e22\u5931<\/li>\n<\/ul>\n<h5>2. \u6c60\u5316\u64cd\u4f5c&#xff1a;<\/h5>\n<p>nn<span class=\"token punctuation\">.<\/span>MaxPool2d<span class=\"token punctuation\">(<\/span><span class=\"token number\">2<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token number\">2<\/span><span class=\"token punctuation\">)<\/span>  <span class=\"token comment\"># 2\u00d72\u6700\u5927\u6c60\u5316&#xff0c;\u6b65\u957f2<\/span><\/p>\n<h6>\u6570\u5b66\u539f\u7406\u4e0e\u63a8\u5bfc&#xff1a;<\/h6>\n<p>\u6c60\u5316\u64cd\u4f5c\u662f\u4e00\u79cd\u4e0b\u91c7\u6837\u6280\u672f&#xff0c;\u7528\u4e8e\u51cf\u5c11\u7279\u5f81\u56fe\u7684\u7a7a\u95f4\u5c3a\u5bf8\u548c\u53c2\u6570\u6570\u91cf&#xff0c;\u540c\u65f6\u589e\u5f3a\u7279\u5f81\u7684\u4e0d\u53d8\u6027\u3002<\/p>\n<p>\u6700\u5927\u6c60\u5316\u516c\u5f0f&#xff1a; <span class=\"katex--display\"><span class=\"katex-display\"><span class=\"katex\"><span class=\"katex-mathml\"><\/p>\n<p>         (<\/p>\n<p>          P<\/p>\n<p>           m<\/p>\n<p>           a<\/p>\n<p>           x<\/p>\n<p>         g<\/p>\n<p>         )<\/p>\n<p>         (<\/p>\n<p>         i<\/p>\n<p>         ,<\/p>\n<p>         j<\/p>\n<p>         ,<\/p>\n<p>         c<\/p>\n<p>         )<\/p>\n<p>         &#061;<\/p>\n<p>         m<\/p>\n<p>         a<\/p>\n<p>          x<\/p>\n<p>           m<\/p>\n<p>           &#061;<\/p>\n<p>           0<\/p>\n<p>          1<\/p>\n<p>         m<\/p>\n<p>         a<\/p>\n<p>          x<\/p>\n<p>           n<\/p>\n<p>           &#061;<\/p>\n<p>           0<\/p>\n<p>          1<\/p>\n<p>         g<\/p>\n<p>         (<\/p>\n<p>         i<\/p>\n<p>         \u2217<\/p>\n<p>         2<\/p>\n<p>         &#043;<\/p>\n<p>         m<\/p>\n<p>         ,<\/p>\n<p>         j<\/p>\n<p>         \u2217<\/p>\n<p>         2<\/p>\n<p>         &#043;<\/p>\n<p>         n<\/p>\n<p>         ,<\/p>\n<p>         c<\/p>\n<p>         )<\/p>\n<p>         (P_{max} g)(i, j, c) &#061; max_{m&#061;0}^{1} max_{n&#061;0}^{1} g(i*2&#043;m, j*2&#043;n, c) <\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 1em;vertical-align: -0.25em\"><\/span><span class=\"mopen\">(<\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.1389em\">P<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.1514em\"><span class=\"\" style=\"top: -2.55em;margin-left: -0.1389em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mathnormal mtight\">ma<\/span><span class=\"mord mathnormal mtight\">x<\/span><\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0359em\">g<\/span><span class=\"mclose\">)<\/span><span class=\"mopen\">(<\/span><span class=\"mord mathnormal\">i<\/span><span class=\"mpunct\">,<\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0572em\">j<\/span><span class=\"mpunct\">,<\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord mathnormal\">c<\/span><span class=\"mclose\">)<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 1.1141em;vertical-align: -0.25em\"><\/span><span class=\"mord mathnormal\">ma<\/span><span class=\"mord\"><span class=\"mord mathnormal\">x<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.8641em\"><span class=\"\" style=\"top: -2.453em;margin-left: 0em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mathnormal mtight\">m<\/span><span class=\"mrel mtight\">&#061;<\/span><span class=\"mord mtight\">0<\/span><\/span><\/span><\/span><span class=\"\" style=\"top: -3.113em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mtight\">1<\/span><\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.247em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mord mathnormal\">ma<\/span><span class=\"mord\"><span class=\"mord mathnormal\">x<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.8641em\"><span class=\"\" style=\"top: -2.453em;margin-left: 0em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mathnormal mtight\">n<\/span><span class=\"mrel mtight\">&#061;<\/span><span class=\"mord mtight\">0<\/span><\/span><\/span><\/span><span class=\"\" style=\"top: -3.113em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mtight\">1<\/span><\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.247em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0359em\">g<\/span><span class=\"mopen\">(<\/span><span class=\"mord mathnormal\">i<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u2217<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.7278em;vertical-align: -0.0833em\"><\/span><span class=\"mord\">2<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">&#043;<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.854em;vertical-align: -0.1944em\"><\/span><span class=\"mord mathnormal\">m<\/span><span class=\"mpunct\">,<\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0572em\">j<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u2217<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.7278em;vertical-align: -0.0833em\"><\/span><span class=\"mord\">2<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">&#043;<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 1em;vertical-align: -0.25em\"><\/span><span class=\"mord mathnormal\">n<\/span><span class=\"mpunct\">,<\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord mathnormal\">c<\/span><span class=\"mclose\">)<\/span><\/span><\/span><\/span><\/span><\/span><\/p>\n<p>\u5e73\u5747\u6c60\u5316\u516c\u5f0f&#xff1a; <span class=\"katex--display\"><span class=\"katex-display\"><span class=\"katex\"><span class=\"katex-mathml\"><\/p>\n<p>         (<\/p>\n<p>          P<\/p>\n<p>           a<\/p>\n<p>           v<\/p>\n<p>           g<\/p>\n<p>         g<\/p>\n<p>         )<\/p>\n<p>         (<\/p>\n<p>         i<\/p>\n<p>         ,<\/p>\n<p>         j<\/p>\n<p>         ,<\/p>\n<p>         c<\/p>\n<p>         )<\/p>\n<p>         &#061;<\/p>\n<p>          1<\/p>\n<p>          4<\/p>\n<p>          \u2211<\/p>\n<p>           m<\/p>\n<p>           &#061;<\/p>\n<p>           0<\/p>\n<p>          1<\/p>\n<p>          \u2211<\/p>\n<p>           n<\/p>\n<p>           &#061;<\/p>\n<p>           0<\/p>\n<p>          1<\/p>\n<p>         g<\/p>\n<p>         (<\/p>\n<p>         i<\/p>\n<p>         \u2217<\/p>\n<p>         2<\/p>\n<p>         &#043;<\/p>\n<p>         m<\/p>\n<p>         ,<\/p>\n<p>         j<\/p>\n<p>         \u2217<\/p>\n<p>         2<\/p>\n<p>         &#043;<\/p>\n<p>         n<\/p>\n<p>         ,<\/p>\n<p>         c<\/p>\n<p>         )<\/p>\n<p>         (P_{avg} g)(i, j, c) &#061; \\\\frac{1}{4} \\\\sum_{m&#061;0}^{1} \\\\sum_{n&#061;0}^{1} g(i*2&#043;m, j*2&#043;n, c) <\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 1.0361em;vertical-align: -0.2861em\"><\/span><span class=\"mopen\">(<\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.1389em\">P<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.1514em\"><span class=\"\" style=\"top: -2.55em;margin-left: -0.1389em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mathnormal mtight\">a<\/span><span class=\"mord mathnormal mtight\" style=\"margin-right: 0.0359em\">vg<\/span><\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.2861em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0359em\">g<\/span><span class=\"mclose\">)<\/span><span class=\"mopen\">(<\/span><span class=\"mord mathnormal\">i<\/span><span class=\"mpunct\">,<\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0572em\">j<\/span><span class=\"mpunct\">,<\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord mathnormal\">c<\/span><span class=\"mclose\">)<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 3.0682em;vertical-align: -1.2671em\"><\/span><span class=\"mord\"><span class=\"mopen nulldelimiter\"><\/span><span class=\"mfrac\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.3214em\"><span class=\"\" style=\"top: -2.314em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"mord\"><span class=\"mord\">4<\/span><\/span><\/span><span class=\"\" style=\"top: -3.23em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"frac-line\" style=\"border-bottom-width: 0.04em\"><\/span><\/span><span class=\"\" style=\"top: -3.677em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"mord\"><span class=\"mord\">1<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.686em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><span class=\"mclose nulldelimiter\"><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mop op-limits\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.8011em\"><span class=\"\" style=\"top: -1.8829em;margin-left: 0em\"><span class=\"pstrut\" style=\"height: 3.05em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mathnormal mtight\">m<\/span><span class=\"mrel mtight\">&#061;<\/span><span class=\"mord mtight\">0<\/span><\/span><\/span><\/span><span class=\"\" style=\"top: -3.05em\"><span class=\"pstrut\" style=\"height: 3.05em\"><\/span><span class=\"\"><span class=\"mop op-symbol large-op\">\u2211<\/span><\/span><\/span><span class=\"\" style=\"top: -4.3em;margin-left: 0em\"><span class=\"pstrut\" style=\"height: 3.05em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mtight\">1<\/span><\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.2671em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mop op-limits\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.8011em\"><span class=\"\" style=\"top: -1.8829em;margin-left: 0em\"><span class=\"pstrut\" style=\"height: 3.05em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mathnormal mtight\">n<\/span><span class=\"mrel mtight\">&#061;<\/span><span class=\"mord mtight\">0<\/span><\/span><\/span><\/span><span class=\"\" style=\"top: -3.05em\"><span class=\"pstrut\" style=\"height: 3.05em\"><\/span><span class=\"\"><span class=\"mop op-symbol large-op\">\u2211<\/span><\/span><\/span><span class=\"\" style=\"top: -4.3em;margin-left: 0em\"><span class=\"pstrut\" style=\"height: 3.05em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mtight\">1<\/span><\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.2671em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0359em\">g<\/span><span class=\"mopen\">(<\/span><span class=\"mord mathnormal\">i<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u2217<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.7278em;vertical-align: -0.0833em\"><\/span><span class=\"mord\">2<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">&#043;<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.854em;vertical-align: -0.1944em\"><\/span><span class=\"mord mathnormal\">m<\/span><span class=\"mpunct\">,<\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0572em\">j<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u2217<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.7278em;vertical-align: -0.0833em\"><\/span><span class=\"mord\">2<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">&#043;<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 1em;vertical-align: -0.25em\"><\/span><span class=\"mord mathnormal\">n<\/span><span class=\"mpunct\">,<\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord mathnormal\">c<\/span><span class=\"mclose\">)<\/span><\/span><\/span><\/span><\/span><\/span><\/p>\n<h6>\u4ee3\u7801\u8bbe\u8ba1\u601d\u8def&#xff1a;<\/h6>\n<ul>\n<li>\u9009\u62e9\u6700\u5927\u6c60\u5316&#xff1a;\u66f4\u597d\u5730\u4fdd\u7559\u7eb9\u7406\u548c\u8fb9\u7f18\u7279\u5f81&#xff0c;\u9002\u5408\u56fe\u50cf\u5206\u7c7b\u4efb\u52a1<\/li>\n<li>2&#215;2\u6c60\u5316\u6838&#xff1a;\u5e38\u7528\u7684\u6c60\u5316\u5c3a\u5bf8&#xff0c;\u5728\u964d\u7ef4\u548c\u4fe1\u606f\u4fdd\u7559\u4e4b\u95f4\u53d6\u5f97\u5e73\u8861<\/li>\n<li>\u6b65\u957f2&#xff1a;\u4f7f\u7279\u5f81\u56fe\u5c3a\u5bf8\u51cf\u534a&#xff0c;\u6709\u6548\u51cf\u5c11\u8ba1\u7b97\u91cf<\/li>\n<\/ul>\n<h6>\u4e0e\u9879\u76ee\u7684\u7ed3\u5408&#xff1a;<\/h6>\n<p>\u8fde\u7eed3\u6b21\u6c60\u5316\u64cd\u4f5c\u5c06\u8f93\u516532&#215;32\u7684\u56fe\u50cf\u9010\u6b65\u964d\u7ef4\u4e3a4&#215;4&#xff1a;<\/p>\n<ul>\n<li>\u7b2c1\u6b21\u6c60\u5316&#xff1a;32&#215;32 \u2192 16&#215;16<\/li>\n<li>\u7b2c2\u6b21\u6c60\u5316&#xff1a;16&#215;16 \u2192 8&#215;8<\/li>\n<li>\u7b2c3\u6b21\u6c60\u5316&#xff1a;8&#215;8 \u2192 4&#215;4<\/li>\n<\/ul>\n<h5>4.3.3 \u5168\u8fde\u63a5\u5c42&#xff1a;\u4ece\u7279\u5f81\u5230\u5206\u7c7b\u7684\u6620\u5c04<\/h5>\n<p>nn<span class=\"token punctuation\">.<\/span>Linear<span class=\"token punctuation\">(<\/span><span class=\"token number\">128<\/span> <span class=\"token operator\">*<\/span> <span class=\"token number\">4<\/span> <span class=\"token operator\">*<\/span> <span class=\"token number\">4<\/span><span class=\"token punctuation\">,<\/span> <span class=\"token number\">512<\/span><span class=\"token punctuation\">)<\/span>  <span class=\"token comment\"># \u8f93\u5165\u7ef4\u5ea62048&#xff0c;\u8f93\u51fa\u7ef4\u5ea6512<\/span><\/p>\n<h6>\u6570\u5b66\u539f\u7406\u4e0e\u63a8\u5bfc&#xff1a;<\/h6>\n<p>\u5168\u8fde\u63a5\u5c42\u5b9e\u73b0\u4e86\u7ebf\u6027\u53d8\u6362&#xff0c;\u5c06\u5377\u79ef\u5c42\u63d0\u53d6\u7684\u7279\u5f81\u6620\u5c04\u5230\u7c7b\u522b\u7a7a\u95f4\u3002<\/p>\n<p>\u7ebf\u6027\u53d8\u6362\u516c\u5f0f&#xff1a; <span class=\"katex--display\"><span class=\"katex-display\"><span class=\"katex\"><span class=\"katex-mathml\"><\/p>\n<p>         y<\/p>\n<p>         &#061;<\/p>\n<p>         W<\/p>\n<p>         x<\/p>\n<p>         &#043;<\/p>\n<p>         b<\/p>\n<p>         y &#061; Wx &#043; b <\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.625em;vertical-align: -0.1944em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0359em\">y<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.7667em;vertical-align: -0.0833em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.1389em\">W<\/span><span class=\"mord mathnormal\">x<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">&#043;<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.6944em\"><\/span><span class=\"mord mathnormal\">b<\/span><\/span><\/span><\/span><\/span><\/span><\/p>\n<p>\u5176\u4e2d&#xff1a;<\/p>\n<ul>\n<li><span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">\n<p>         x<\/p>\n<p>        x<\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.4306em\"><\/span><span class=\"mord mathnormal\">x<\/span><\/span><\/span><\/span><\/span>&#xff1a;\u8f93\u5165\u5411\u91cf&#xff0c;\u7ef4\u5ea6\u4e3a<span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\"><\/p>\n<p>         128<\/p>\n<p>         \u2217<\/p>\n<p>         4<\/p>\n<p>         \u2217<\/p>\n<p>         4<\/p>\n<p>         &#061;<\/p>\n<p>         2048<\/p>\n<p>        128*4*4&#061;2048<\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.6444em\"><\/span><span class=\"mord\">128<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u2217<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.6444em\"><\/span><span class=\"mord\">4<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u2217<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.6444em\"><\/span><span class=\"mord\">4<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.6444em\"><\/span><span class=\"mord\">2048<\/span><\/span><\/span><\/span><\/span><\/li>\n<li><span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">\n<p>         W<\/p>\n<p>        W<\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.6833em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.1389em\">W<\/span><\/span><\/span><\/span><\/span>&#xff1a;\u6743\u91cd\u77e9\u9635&#xff0c;\u7ef4\u5ea6\u4e3a<span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\"><\/p>\n<p>         512<\/p>\n<p>         \u00d7<\/p>\n<p>         2048<\/p>\n<p>        512\u00d72048<\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.7278em;vertical-align: -0.0833em\"><\/span><span class=\"mord\">512<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u00d7<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.6444em\"><\/span><span class=\"mord\">2048<\/span><\/span><\/span><\/span><\/span><\/li>\n<li><span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">\n<p>         b<\/p>\n<p>        b<\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.6944em\"><\/span><span class=\"mord mathnormal\">b<\/span><\/span><\/span><\/span><\/span>&#xff1a;\u504f\u7f6e\u5411\u91cf&#xff0c;\u7ef4\u5ea6\u4e3a<span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\"><\/p>\n<p>         512<\/p>\n<p>        512<\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.6444em\"><\/span><span class=\"mord\">512<\/span><\/span><\/span><\/span><\/span><\/li>\n<li><span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">\n<p>         y<\/p>\n<p>        y<\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.625em;vertical-align: -0.1944em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0359em\">y<\/span><\/span><\/span><\/span><\/span>&#xff1a;\u8f93\u51fa\u5411\u91cf&#xff0c;\u7ef4\u5ea6\u4e3a<span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\"><\/p>\n<p>         512<\/p>\n<p>        512<\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.6444em\"><\/span><span class=\"mord\">512<\/span><\/span><\/span><\/span><\/span><\/li>\n<\/ul>\n<p>\u53c2\u6570\u6570\u91cf\u8ba1\u7b97&#xff1a; <span class=\"katex--display\"><span class=\"katex-display\"><span class=\"katex\"><span class=\"katex-mathml\"><\/p>\n<p>         P<\/p>\n<p>         a<\/p>\n<p>         r<\/p>\n<p>         a<\/p>\n<p>         m<\/p>\n<p>         e<\/p>\n<p>         t<\/p>\n<p>         e<\/p>\n<p>         r<\/p>\n<p>         s<\/p>\n<p>         &#061;<\/p>\n<p>         (<\/p>\n<p>         I<\/p>\n<p>         n<\/p>\n<p>         p<\/p>\n<p>         u<\/p>\n<p>         t<\/p>\n<p>         _<\/p>\n<p>         d<\/p>\n<p>         i<\/p>\n<p>         m<\/p>\n<p>         \u2217<\/p>\n<p>         O<\/p>\n<p>         u<\/p>\n<p>         t<\/p>\n<p>         p<\/p>\n<p>         u<\/p>\n<p>         t<\/p>\n<p>         _<\/p>\n<p>         d<\/p>\n<p>         i<\/p>\n<p>         m<\/p>\n<p>         )<\/p>\n<p>         &#043;<\/p>\n<p>         O<\/p>\n<p>         u<\/p>\n<p>         t<\/p>\n<p>         p<\/p>\n<p>         u<\/p>\n<p>         t<\/p>\n<p>         _<\/p>\n<p>         d<\/p>\n<p>         i<\/p>\n<p>         m<\/p>\n<p>         Parameters &#061; (Input\\\\_dim * Output\\\\_dim) &#043; Output\\\\_dim <\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.6833em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.1389em\">P<\/span><span class=\"mord mathnormal\">a<\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0278em\">r<\/span><span class=\"mord mathnormal\">am<\/span><span class=\"mord mathnormal\">e<\/span><span class=\"mord mathnormal\">t<\/span><span class=\"mord mathnormal\">ers<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 1.06em;vertical-align: -0.31em\"><\/span><span class=\"mopen\">(<\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0785em\">I<\/span><span class=\"mord mathnormal\">n<\/span><span class=\"mord mathnormal\">p<\/span><span class=\"mord mathnormal\">u<\/span><span class=\"mord mathnormal\">t<\/span><span class=\"mord\" style=\"margin-right: 0.0278em\">_<\/span><span class=\"mord mathnormal\">d<\/span><span class=\"mord mathnormal\">im<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u2217<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 1.06em;vertical-align: -0.31em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0278em\">O<\/span><span class=\"mord mathnormal\">u<\/span><span class=\"mord mathnormal\">tp<\/span><span class=\"mord mathnormal\">u<\/span><span class=\"mord mathnormal\">t<\/span><span class=\"mord\" style=\"margin-right: 0.0278em\">_<\/span><span class=\"mord mathnormal\">d<\/span><span class=\"mord mathnormal\">im<\/span><span class=\"mclose\">)<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">&#043;<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 1.0044em;vertical-align: -0.31em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0278em\">O<\/span><span class=\"mord mathnormal\">u<\/span><span class=\"mord mathnormal\">tp<\/span><span class=\"mord mathnormal\">u<\/span><span class=\"mord mathnormal\">t<\/span><span class=\"mord\" style=\"margin-right: 0.0278em\">_<\/span><span class=\"mord mathnormal\">d<\/span><span class=\"mord mathnormal\">im<\/span><\/span><\/span><\/span><\/span><\/span><\/p>\n<p>\u5bf9\u4e8e\u7b2c\u4e00\u4e2a\u5168\u8fde\u63a5\u5c42&#xff1a; <span class=\"katex--display\"><span class=\"katex-display\"><span class=\"katex\"><span class=\"katex-mathml\"><\/p>\n<p>         P<\/p>\n<p>         a<\/p>\n<p>         r<\/p>\n<p>         a<\/p>\n<p>         m<\/p>\n<p>         e<\/p>\n<p>         t<\/p>\n<p>         e<\/p>\n<p>         r<\/p>\n<p>         s<\/p>\n<p>         &#061;<\/p>\n<p>         (<\/p>\n<p>         2048<\/p>\n<p>         \u2217<\/p>\n<p>         512<\/p>\n<p>         )<\/p>\n<p>         &#043;<\/p>\n<p>         512<\/p>\n<p>         &#061;<\/p>\n<p>         1<\/p>\n<p>         ,<\/p>\n<p>         049<\/p>\n<p>         ,<\/p>\n<p>         088<\/p>\n<p>         Parameters &#061; (2048 * 512) &#043; 512 &#061; 1,049,088 <\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.6833em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.1389em\">P<\/span><span class=\"mord mathnormal\">a<\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0278em\">r<\/span><span class=\"mord mathnormal\">am<\/span><span class=\"mord mathnormal\">e<\/span><span class=\"mord mathnormal\">t<\/span><span class=\"mord mathnormal\">ers<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 1em;vertical-align: -0.25em\"><\/span><span class=\"mopen\">(<\/span><span class=\"mord\">2048<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u2217<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 1em;vertical-align: -0.25em\"><\/span><span class=\"mord\">512<\/span><span class=\"mclose\">)<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">&#043;<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.6444em\"><\/span><span class=\"mord\">512<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.8389em;vertical-align: -0.1944em\"><\/span><span class=\"mord\">1<\/span><span class=\"mpunct\">,<\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord\">049<\/span><span class=\"mpunct\">,<\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord\">088<\/span><\/span><\/span><\/span><\/span><\/span><\/p>\n<h6>\u4ee3\u7801\u8bbe\u8ba1\u601d\u8def&#xff1a;<\/h6>\n<ul>\n<li>\u9010\u6b65\u51cf\u5c11\u7ef4\u5ea6&#xff1a;2048 \u2192 512 \u2192 256 \u2192 128 \u2192 10&#xff0c;\u5f62\u6210&#034;\u6f0f\u6597&#034;\u5f62\u72b6<\/li>\n<li>\u6df1\u5ea6\u7ed3\u6784&#xff1a;\u591a\u5c42\u5168\u8fde\u63a5\u5c42\u80fd\u591f\u5b66\u4e60\u66f4\u590d\u6742\u7684\u975e\u7ebf\u6027\u6620\u5c04<\/li>\n<li>\u7ef4\u5ea6\u5339\u914d&#xff1a;\u786e\u4fdd\u5377\u79ef\u5c42\u8f93\u51fa\u7684\u7279\u5f81\u56fe\u5c55\u5e73\u540e\u4e0e\u5168\u8fde\u63a5\u5c42\u8f93\u5165\u7ef4\u5ea6\u4e00\u81f4<\/li>\n<\/ul>\n<h5>4. \u6fc0\u6d3b\u51fd\u6570&#xff1a;<\/h5>\n<p>nn<span class=\"token punctuation\">.<\/span>ReLU<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span>  <span class=\"token comment\"># ReLU\u6fc0\u6d3b\u51fd\u6570<\/span><\/p>\n<h6>\u6570\u5b66\u539f\u7406\u4e0e\u63a8\u5bfc&#xff1a;<\/h6>\n<p>\u6fc0\u6d3b\u51fd\u6570\u5f15\u5165\u975e\u7ebf\u6027&#xff0c;\u4f7f\u7f51\u7edc\u80fd\u591f\u5b66\u4e60\u590d\u6742\u7684\u975e\u7ebf\u6027\u51fd\u6570\u3002<\/p>\n<p>ReLU\u51fd\u6570\u516c\u5f0f&#xff1a; <span class=\"katex--display\"><span class=\"katex-display\"><span class=\"katex\"><span class=\"katex-mathml\"><\/p>\n<p>         f<\/p>\n<p>         (<\/p>\n<p>         z<\/p>\n<p>         )<\/p>\n<p>         &#061;<\/p>\n<p>         m<\/p>\n<p>         a<\/p>\n<p>         x<\/p>\n<p>         (<\/p>\n<p>         0<\/p>\n<p>         ,<\/p>\n<p>         z<\/p>\n<p>         )<\/p>\n<p>         f(z) &#061; max(0, z) <\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 1em;vertical-align: -0.25em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.1076em\">f<\/span><span class=\"mopen\">(<\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.044em\">z<\/span><span class=\"mclose\">)<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 1em;vertical-align: -0.25em\"><\/span><span class=\"mord mathnormal\">ma<\/span><span class=\"mord mathnormal\">x<\/span><span class=\"mopen\">(<\/span><span class=\"mord\">0<\/span><span class=\"mpunct\">,<\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.044em\">z<\/span><span class=\"mclose\">)<\/span><\/span><\/span><\/span><\/span><\/span><\/p>\n<p>\u5bfc\u6570&#xff1a; <span class=\"katex--display\"><span class=\"katex-display\"><span class=\"katex\"><span class=\"katex-mathml\"><\/p>\n<p>          f<\/p>\n<p>          \u2032<\/p>\n<p>         (<\/p>\n<p>         z<\/p>\n<p>         )<\/p>\n<p>         &#061;<\/p>\n<p>          {<\/p>\n<p>               1<\/p>\n<p>               ,<\/p>\n<p>               z<\/p>\n<p>               &gt;<\/p>\n<p>               0<\/p>\n<p>               0<\/p>\n<p>               ,<\/p>\n<p>               z<\/p>\n<p>               \u2264<\/p>\n<p>               0<\/p>\n<p>         f&#039;(z) &#061; \\\\begin{cases} 1, &amp; z &gt; 0 \\\\\\\\ 0, &amp; z \\\\leq 0 \\\\end{cases} <\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 1.0519em;vertical-align: -0.25em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.1076em\">f<\/span><span class=\"msupsub\"><span class=\"vlist-t\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.8019em\"><span class=\"\" style=\"top: -3.113em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mtight\">\u2032<\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mopen\">(<\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.044em\">z<\/span><span class=\"mclose\">)<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 3em;vertical-align: -1.25em\"><\/span><span class=\"minner\"><span class=\"mopen delimcenter\" style=\"top: 0em\"><span class=\"delimsizing size4\">{<\/span><\/span><span class=\"mord\"><span class=\"mtable\"><span class=\"col-align-l\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.69em\"><span class=\"\" style=\"top: -3.69em\"><span class=\"pstrut\" style=\"height: 3.008em\"><\/span><span class=\"mord\"><span class=\"mord\">1<\/span><span class=\"mpunct\">,<\/span><\/span><\/span><span class=\"\" style=\"top: -2.25em\"><span class=\"pstrut\" style=\"height: 3.008em\"><\/span><span class=\"mord\"><span class=\"mord\">0<\/span><span class=\"mpunct\">,<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.19em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><span class=\"arraycolsep\" style=\"width: 1em\"><\/span><span class=\"col-align-l\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.69em\"><span class=\"\" style=\"top: -3.69em\"><span class=\"pstrut\" style=\"height: 3.008em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.044em\">z<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&gt;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mord\">0<\/span><\/span><\/span><span class=\"\" style=\"top: -2.25em\"><span class=\"pstrut\" style=\"height: 3.008em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.044em\">z<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">\u2264<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mord\">0<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.19em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mclose nulldelimiter\"><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/p>\n<h6>\u4ee3\u7801\u8bbe\u8ba1\u601d\u8def&#xff1a;<\/h6>\n<ul>\n<li>\u9009\u62e9ReLU&#xff1a;\u89e3\u51b3\u4e86sigmoid\u548ctanh\u7684\u68af\u5ea6\u6d88\u5931\u95ee\u9898&#xff0c;\u8ba1\u7b97\u6548\u7387\u9ad8<\/li>\n<li>\u6bcf\u4e00\u5c42\u540e\u4f7f\u7528&#xff1a;\u5728\u5377\u79ef\u5c42\u548c\u5168\u8fde\u63a5\u5c42\u540e\u90fd\u4f7f\u7528ReLU&#xff0c;\u5f15\u5165\u975e\u7ebf\u6027<\/li>\n<li>\u907f\u514d\u6b7b\u4ea1\u795e\u7ecf\u5143&#xff1a;\u901a\u8fc7\u5408\u7406\u7684\u6743\u91cd\u521d\u59cb\u5316&#xff08;\u5982He\u521d\u59cb\u5316&#xff09;\u51cf\u5c11ReLU\u795e\u7ecf\u5143\u6b7b\u4ea1\u7684\u6982\u7387<\/li>\n<\/ul>\n<h6>\u4e0e\u9879\u76ee\u7684\u7ed3\u5408&#xff1a;<\/h6>\n<p>\u5728\u672c\u9879\u76ee\u4e2d&#xff0c;ReLU\u6fc0\u6d3b\u51fd\u6570\u5e94\u7528\u4e8e&#xff1a;<\/p>\n<ul>\n<li>\u6bcf\u4e2a\u5377\u79ef\u5c42\u4e4b\u540e<\/li>\n<li>\u6bcf\u4e2a\u5168\u8fde\u63a5\u5c42\u4e4b\u540e&#xff08;\u9664\u4e86\u8f93\u51fa\u5c42&#xff09;<\/li>\n<\/ul>\n<h5>4.3.5 \u635f\u5931\u4e0e\u4f18\u5316&#xff1a;\u6a21\u578b\u5b66\u4e60\u7684\u6838\u5fc3\u673a\u5236<\/h5>\n<p>criterion <span class=\"token operator\">&#061;<\/span> nn<span class=\"token punctuation\">.<\/span>CrossEntropyLoss<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span>  <span class=\"token comment\"># \u4ea4\u53c9\u71b5\u635f\u5931<\/span><br \/>\noptimizer <span class=\"token operator\">&#061;<\/span> optim<span class=\"token punctuation\">.<\/span>Adam<span class=\"token punctuation\">(<\/span>model<span class=\"token punctuation\">.<\/span>parameters<span class=\"token punctuation\">(<\/span><span class=\"token punctuation\">)<\/span><span class=\"token punctuation\">,<\/span> lr<span class=\"token operator\">&#061;<\/span><span class=\"token number\">1e-3<\/span><span class=\"token punctuation\">)<\/span>  <span class=\"token comment\"># Adam\u4f18\u5316\u5668<\/span><\/p>\n<h6>\u4ea4\u53c9\u71b5\u635f\u5931\u7684\u6570\u5b66\u63a8\u5bfc&#xff1a;<\/h6>\n<p>\u5bf9\u4e8e\u591a\u5206\u7c7b\u95ee\u9898&#xff0c;\u5047\u8bbe\u6a21\u578b\u8f93\u51fa\u4e3a<span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\"><\/p>\n<p>         z<\/p>\n<p>         1<\/p>\n<p>        ,<\/p>\n<p>         z<\/p>\n<p>         2<\/p>\n<p>        ,<\/p>\n<p>        .<\/p>\n<p>        .<\/p>\n<p>        .<\/p>\n<p>        ,<\/p>\n<p>         z<\/p>\n<p>         C<\/p>\n<p>       z_1, z_2, &#8230;, z_C<\/p>\n<p>    <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.625em;vertical-align: -0.1944em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.044em\">z<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.3011em\"><span class=\"\" style=\"top: -2.55em;margin-left: -0.044em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\">1<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mpunct\">,<\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.044em\">z<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.3011em\"><span class=\"\" style=\"top: -2.55em;margin-left: -0.044em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\">2<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mpunct\">,<\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord\">&#8230;<\/span><span class=\"mpunct\">,<\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.044em\">z<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.3283em\"><span class=\"\" style=\"top: -2.55em;margin-left: -0.044em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mathnormal mtight\" style=\"margin-right: 0.0715em\">C<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span>&#xff08;<span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\"><\/p>\n<p>        C<\/p>\n<p>       C<\/p>\n<p>    <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.6833em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0715em\">C<\/span><\/span><\/span><\/span><\/span>\u4e3a\u7c7b\u522b\u6570&#xff09;&#xff0c;\u9996\u5148\u4f7f\u7528softmax\u51fd\u6570\u5c06\u5176\u8f6c\u6362\u4e3a\u6982\u7387&#xff1a;<\/p>\n<p><span class=\"katex--display\"><span class=\"katex-display\"><span class=\"katex\"><span class=\"katex-mathml\"><\/p>\n<p>           y<\/p>\n<p>           ^<\/p>\n<p>          i<\/p>\n<p>         &#061;<\/p>\n<p>         s<\/p>\n<p>         o<\/p>\n<p>         f<\/p>\n<p>         t<\/p>\n<p>         m<\/p>\n<p>         a<\/p>\n<p>         x<\/p>\n<p>         (<\/p>\n<p>          z<\/p>\n<p>          i<\/p>\n<p>         )<\/p>\n<p>         &#061;<\/p>\n<p>           e<\/p>\n<p>            z<\/p>\n<p>            i<\/p>\n<p>            \u2211<\/p>\n<p>             j<\/p>\n<p>             &#061;<\/p>\n<p>             1<\/p>\n<p>            C<\/p>\n<p>            e<\/p>\n<p>             z<\/p>\n<p>             j<\/p>\n<p>         \\\\hat{y}_i &#061; softmax(z_i) &#061; \\\\frac{e^{z_i}}{\\\\sum_{j&#061;1}^{C} e^{z_j}} <\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.8889em;vertical-align: -0.1944em\"><\/span><span class=\"mord\"><span class=\"mord accent\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.6944em\"><span class=\"\" style=\"top: -3em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0359em\">y<\/span><\/span><span class=\"\" style=\"top: -3em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"accent-body\" style=\"left: -0.1944em\"><span class=\"mord\">^<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.1944em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.3117em\"><span class=\"\" style=\"top: -2.55em;margin-left: -0.0359em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mathnormal mtight\">i<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 1em;vertical-align: -0.25em\"><\/span><span class=\"mord mathnormal\">so<\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.1076em\">f<\/span><span class=\"mord mathnormal\">t<\/span><span class=\"mord mathnormal\">ma<\/span><span class=\"mord mathnormal\">x<\/span><span class=\"mopen\">(<\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.044em\">z<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.3117em\"><span class=\"\" style=\"top: -2.55em;margin-left: -0.044em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mathnormal mtight\">i<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mclose\">)<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 2.6484em;vertical-align: -1.307em\"><\/span><span class=\"mord\"><span class=\"mopen nulldelimiter\"><\/span><span class=\"mfrac\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.3414em\"><span class=\"\" style=\"top: -2.1288em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"mord\"><span class=\"mop\"><span class=\"mop op-symbol small-op\" style=\"position: relative;top: 0em\">\u2211<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.9812em\"><span class=\"\" style=\"top: -2.4003em;margin-left: 0em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mathnormal mtight\" style=\"margin-right: 0.0572em\">j<\/span><span class=\"mrel mtight\">&#061;<\/span><span class=\"mord mtight\">1<\/span><\/span><\/span><\/span><span class=\"\" style=\"top: -3.2029em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mathnormal mtight\" style=\"margin-right: 0.0715em\">C<\/span><\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.4358em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\">e<\/span><span class=\"msupsub\"><span class=\"vlist-t\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.6065em\"><span class=\"\" style=\"top: -3.0051em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mtight\"><span class=\"mord mathnormal mtight\" style=\"margin-right: 0.044em\">z<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.3281em\"><span class=\"\" style=\"top: -2.357em;margin-left: -0.044em;margin-right: 0.0714em\"><span class=\"pstrut\" style=\"height: 2.5em\"><\/span><span class=\"sizing reset-size3 size1 mtight\"><span class=\"mord mathnormal mtight\" style=\"margin-right: 0.0572em\">j<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.2819em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"\" style=\"top: -3.23em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"frac-line\" style=\"border-bottom-width: 0.04em\"><\/span><\/span><span class=\"\" style=\"top: -3.677em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"mord\"><span class=\"mord\"><span class=\"mord mathnormal\">e<\/span><span class=\"msupsub\"><span class=\"vlist-t\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.6644em\"><span class=\"\" style=\"top: -3.063em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mtight\"><span class=\"mord mathnormal mtight\" style=\"margin-right: 0.044em\">z<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.3281em\"><span class=\"\" style=\"top: -2.357em;margin-left: -0.044em;margin-right: 0.0714em\"><span class=\"pstrut\" style=\"height: 2.5em\"><\/span><span class=\"sizing reset-size3 size1 mtight\"><span class=\"mord mathnormal mtight\">i<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.143em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.307em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><span class=\"mclose nulldelimiter\"><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/p>\n<p>\u4ea4\u53c9\u71b5\u635f\u5931\u5b9a\u4e49\u4e3a&#xff1a; <span class=\"katex--display\"><span class=\"katex-display\"><span class=\"katex\"><span class=\"katex-mathml\"><\/p>\n<p>         L<\/p>\n<p>         &#061;<\/p>\n<p>         \u2212<\/p>\n<p>          \u2211<\/p>\n<p>           i<\/p>\n<p>           &#061;<\/p>\n<p>           1<\/p>\n<p>          C<\/p>\n<p>          y<\/p>\n<p>          i<\/p>\n<p>         log<\/p>\n<p>         \u2061<\/p>\n<p>         (<\/p>\n<p>           y<\/p>\n<p>           ^<\/p>\n<p>          i<\/p>\n<p>         )<\/p>\n<p>         L &#061; -\\\\sum_{i&#061;1}^{C} y_i \\\\log(\\\\hat{y}_i) <\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.6833em\"><\/span><span class=\"mord mathnormal\">L<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 3.106em;vertical-align: -1.2777em\"><\/span><span class=\"mord\">\u2212<\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mop op-limits\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.8283em\"><span class=\"\" style=\"top: -1.8723em;margin-left: 0em\"><span class=\"pstrut\" style=\"height: 3.05em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mathnormal mtight\">i<\/span><span class=\"mrel mtight\">&#061;<\/span><span class=\"mord mtight\">1<\/span><\/span><\/span><\/span><span class=\"\" style=\"top: -3.05em\"><span class=\"pstrut\" style=\"height: 3.05em\"><\/span><span class=\"\"><span class=\"mop op-symbol large-op\">\u2211<\/span><\/span><\/span><span class=\"\" style=\"top: -4.3em;margin-left: 0em\"><span class=\"pstrut\" style=\"height: 3.05em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mathnormal mtight\" style=\"margin-right: 0.0715em\">C<\/span><\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.2777em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.0359em\">y<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.3117em\"><span class=\"\" style=\"top: -2.55em;margin-left: -0.0359em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mathnormal mtight\">i<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.1667em\"><\/span><span class=\"mop\">lo<span style=\"margin-right: 0.0139em\">g<\/span><\/span><span class=\"mopen\">(<\/span><span class=\"mord\"><span class=\"mord accent\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.6944em\"><span class=\"\" style=\"top: -3em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0359em\">y<\/span><\/span><span class=\"\" style=\"top: -3em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"accent-body\" style=\"left: -0.1944em\"><span class=\"mord\">^<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.1944em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.3117em\"><span class=\"\" style=\"top: -2.55em;margin-left: -0.0359em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mathnormal mtight\">i<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mclose\">)<\/span><\/span><\/span><\/span><\/span><\/span><\/p>\n<p>\u5176\u4e2d<span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\"><\/p>\n<p>         y<\/p>\n<p>         i<\/p>\n<p>       y_i<\/p>\n<p>    <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.625em;vertical-align: -0.1944em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.0359em\">y<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.3117em\"><span class=\"\" style=\"top: -2.55em;margin-left: -0.0359em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mathnormal mtight\">i<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span>\u662f\u771f\u5b9e\u6807\u7b7e\u7684\u72ec\u70ed\u7f16\u7801&#xff08;\u672c\u9879\u76ee\u4e2dCIFAR10\u7684C&#061;10&#xff09;\u3002<\/p>\n<p>PyTorch\u4e2d\u7684nn.CrossEntropyLoss\u5b9e\u9645\u4e0a\u7ed3\u5408\u4e86softmax\u548c\u4ea4\u53c9\u71b5\u635f\u5931&#xff0c;\u76f4\u63a5\u63a5\u53d7\u6a21\u578b\u7684\u539f\u59cb\u8f93\u51fa\u3002<\/p>\n<h6>Adam\u4f18\u5316\u5668\u7684\u539f\u7406&#xff1a;<\/h6>\n<p>Adam\u4f18\u5316\u5668\u7ed3\u5408\u4e86\u52a8\u91cf\u6cd5\u548cRMSProp\u7684\u4f18\u70b9&#xff0c;\u4f7f\u7528\u81ea\u9002\u5e94\u5b66\u4e60\u7387&#xff1a;<\/p>\n<li>\n<p>\u8ba1\u7b97\u68af\u5ea6\u7684\u4e00\u9636\u77e9\u4f30\u8ba1&#xff08;\u52a8\u91cf&#xff09;&#xff1a; <span class=\"katex--display\"><span class=\"katex-display\"><span class=\"katex\"><span class=\"katex-mathml\"><\/p>\n<p>            m<\/p>\n<p>            t<\/p>\n<p>           &#061;<\/p>\n<p>            \u03b2<\/p>\n<p>            1<\/p>\n<p>            m<\/p>\n<p>             t<\/p>\n<p>             \u2212<\/p>\n<p>             1<\/p>\n<p>           &#043;<\/p>\n<p>           (<\/p>\n<p>           1<\/p>\n<p>           \u2212<\/p>\n<p>            \u03b2<\/p>\n<p>            1<\/p>\n<p>           )<\/p>\n<p>            g<\/p>\n<p>            t<\/p>\n<p>           m_t &#061; \\\\beta_1 m_{t-1} &#043; (1 &#8211; \\\\beta_1) g_t <\/p>\n<p>       <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.5806em;vertical-align: -0.15em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\">m<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.2806em\"><span class=\"\" style=\"top: -2.55em;margin-left: 0em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mathnormal mtight\">t<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.9028em;vertical-align: -0.2083em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.0528em\">\u03b2<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.3011em\"><span class=\"\" style=\"top: -2.55em;margin-left: -0.0528em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\">1<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mord\"><span class=\"mord mathnormal\">m<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.3011em\"><span class=\"\" style=\"top: -2.55em;margin-left: 0em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mathnormal mtight\">t<\/span><span class=\"mbin mtight\">\u2212<\/span><span class=\"mord mtight\">1<\/span><\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.2083em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">&#043;<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 1em;vertical-align: -0.25em\"><\/span><span class=\"mopen\">(<\/span><span class=\"mord\">1<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u2212<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 1em;vertical-align: -0.25em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.0528em\">\u03b2<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.3011em\"><span class=\"\" style=\"top: -2.55em;margin-left: -0.0528em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\">1<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mclose\">)<\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.0359em\">g<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.2806em\"><span class=\"\" style=\"top: -2.55em;margin-left: -0.0359em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mathnormal mtight\">t<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/p>\n<\/li>\n<li>\n<p>\u8ba1\u7b97\u68af\u5ea6\u7684\u4e8c\u9636\u77e9\u4f30\u8ba1&#xff08;RMSProp&#xff09;&#xff1a; <span class=\"katex--display\"><span class=\"katex-display\"><span class=\"katex\"><span class=\"katex-mathml\"><\/p>\n<p>            v<\/p>\n<p>            t<\/p>\n<p>           &#061;<\/p>\n<p>            \u03b2<\/p>\n<p>            2<\/p>\n<p>            v<\/p>\n<p>             t<\/p>\n<p>             \u2212<\/p>\n<p>             1<\/p>\n<p>           &#043;<\/p>\n<p>           (<\/p>\n<p>           1<\/p>\n<p>           \u2212<\/p>\n<p>            \u03b2<\/p>\n<p>            2<\/p>\n<p>           )<\/p>\n<p>            g<\/p>\n<p>            t<\/p>\n<p>            2<\/p>\n<p>           v_t &#061; \\\\beta_2 v_{t-1} &#043; (1 &#8211; \\\\beta_2) g_t^2 <\/p>\n<p>       <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.5806em;vertical-align: -0.15em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.0359em\">v<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.2806em\"><span class=\"\" style=\"top: -2.55em;margin-left: -0.0359em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mathnormal mtight\">t<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.9028em;vertical-align: -0.2083em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.0528em\">\u03b2<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.3011em\"><span class=\"\" style=\"top: -2.55em;margin-left: -0.0528em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\">2<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.0359em\">v<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.3011em\"><span class=\"\" style=\"top: -2.55em;margin-left: -0.0359em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mathnormal mtight\">t<\/span><span class=\"mbin mtight\">\u2212<\/span><span class=\"mord mtight\">1<\/span><\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.2083em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">&#043;<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 1em;vertical-align: -0.25em\"><\/span><span class=\"mopen\">(<\/span><span class=\"mord\">1<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u2212<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 1.1141em;vertical-align: -0.25em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.0528em\">\u03b2<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.3011em\"><span class=\"\" style=\"top: -2.55em;margin-left: -0.0528em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\">2<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mclose\">)<\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.0359em\">g<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.8641em\"><span class=\"\" style=\"top: -2.453em;margin-left: -0.0359em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mathnormal mtight\">t<\/span><\/span><\/span><span class=\"\" style=\"top: -3.113em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\">2<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.247em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/p>\n<\/li>\n<li>\n<p>\u504f\u5dee\u4fee\u6b63&#xff1a; <span class=\"katex--display\"><span class=\"katex-display\"><span class=\"katex\"><span class=\"katex-mathml\"><\/p>\n<p>             m<\/p>\n<p>             ^<\/p>\n<p>            t<\/p>\n<p>           &#061;<\/p>\n<p>             m<\/p>\n<p>             t<\/p>\n<p>             1<\/p>\n<p>             \u2212<\/p>\n<p>              \u03b2<\/p>\n<p>              1<\/p>\n<p>              t<\/p>\n<p>           \\\\hat{m}_t &#061; \\\\frac{m_t}{1 &#8211; \\\\beta_1^t} <\/p>\n<p>       <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.8444em;vertical-align: -0.15em\"><\/span><span class=\"mord\"><span class=\"mord accent\"><span class=\"vlist-t\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.6944em\"><span class=\"\" style=\"top: -3em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"mord mathnormal\">m<\/span><\/span><span class=\"\" style=\"top: -3em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"accent-body\" style=\"left: -0.25em\"><span class=\"mord\">^<\/span><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.2806em\"><span class=\"\" style=\"top: -2.55em;margin-left: 0em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mathnormal mtight\">t<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 2.0599em;vertical-align: -0.9523em\"><\/span><span class=\"mord\"><span class=\"mopen nulldelimiter\"><\/span><span class=\"mfrac\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.1076em\"><span class=\"\" style=\"top: -2.314em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"mord\"><span class=\"mord\">1<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u2212<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.0528em\">\u03b2<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.7754em\"><span class=\"\" style=\"top: -2.4337em;margin-left: -0.0528em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\">1<\/span><\/span><\/span><span class=\"\" style=\"top: -3.0448em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mathnormal mtight\">t<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.2663em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"\" style=\"top: -3.23em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"frac-line\" style=\"border-bottom-width: 0.04em\"><\/span><\/span><span class=\"\" style=\"top: -3.677em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"mord\"><span class=\"mord\"><span class=\"mord mathnormal\">m<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.2806em\"><span class=\"\" style=\"top: -2.55em;margin-left: 0em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mathnormal mtight\">t<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.9523em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><span class=\"mclose nulldelimiter\"><\/span><\/span><\/span><\/span><\/span><\/span><\/span> <span class=\"katex--display\"><span class=\"katex-display\"><span class=\"katex\"><span class=\"katex-mathml\"><\/p>\n<p>             v<\/p>\n<p>             ^<\/p>\n<p>            t<\/p>\n<p>           &#061;<\/p>\n<p>             v<\/p>\n<p>             t<\/p>\n<p>             1<\/p>\n<p>             \u2212<\/p>\n<p>              \u03b2<\/p>\n<p>              2<\/p>\n<p>              t<\/p>\n<p>           \\\\hat{v}_t &#061; \\\\frac{v_t}{1 &#8211; \\\\beta_2^t} <\/p>\n<p>       <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.8444em;vertical-align: -0.15em\"><\/span><span class=\"mord\"><span class=\"mord accent\"><span class=\"vlist-t\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.6944em\"><span class=\"\" style=\"top: -3em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0359em\">v<\/span><\/span><span class=\"\" style=\"top: -3em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"accent-body\" style=\"left: -0.2222em\"><span class=\"mord\">^<\/span><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.2806em\"><span class=\"\" style=\"top: -2.55em;margin-left: -0.0359em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mathnormal mtight\">t<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 2.0599em;vertical-align: -0.9523em\"><\/span><span class=\"mord\"><span class=\"mopen nulldelimiter\"><\/span><span class=\"mfrac\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.1076em\"><span class=\"\" style=\"top: -2.314em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"mord\"><span class=\"mord\">1<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u2212<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.0528em\">\u03b2<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.7754em\"><span class=\"\" style=\"top: -2.4337em;margin-left: -0.0528em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\">2<\/span><\/span><\/span><span class=\"\" style=\"top: -3.0448em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mathnormal mtight\">t<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.2663em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"\" style=\"top: -3.23em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"frac-line\" style=\"border-bottom-width: 0.04em\"><\/span><\/span><span class=\"\" style=\"top: -3.677em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"mord\"><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.0359em\">v<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.2806em\"><span class=\"\" style=\"top: -2.55em;margin-left: -0.0359em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mathnormal mtight\">t<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.9523em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><span class=\"mclose nulldelimiter\"><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/p>\n<\/li>\n<li>\n<p>\u53c2\u6570\u66f4\u65b0&#xff1a; <span class=\"katex--display\"><span class=\"katex-display\"><span class=\"katex\"><span class=\"katex-mathml\"><\/p>\n<p>            w<\/p>\n<p>            t<\/p>\n<p>           &#061;<\/p>\n<p>            w<\/p>\n<p>             t<\/p>\n<p>             \u2212<\/p>\n<p>             1<\/p>\n<p>           \u2212<\/p>\n<p>           \u03b7<\/p>\n<p>              m<\/p>\n<p>              ^<\/p>\n<p>             t<\/p>\n<p>                v<\/p>\n<p>                ^<\/p>\n<p>               t<\/p>\n<p>             &#043;<\/p>\n<p>             \u03f5<\/p>\n<p>           w_t &#061; w_{t-1} &#8211; \\\\eta \\\\frac{\\\\hat{m}_t}{\\\\sqrt{\\\\hat{v}_t} &#043; \\\\epsilon} <\/p>\n<p>       <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.5806em;vertical-align: -0.15em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.0269em\">w<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.2806em\"><span class=\"\" style=\"top: -2.55em;margin-left: -0.0269em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mathnormal mtight\">t<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.7917em;vertical-align: -0.2083em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.0269em\">w<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.3011em\"><span class=\"\" style=\"top: -2.55em;margin-left: -0.0269em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mathnormal mtight\">t<\/span><span class=\"mbin mtight\">\u2212<\/span><span class=\"mord mtight\">1<\/span><\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.2083em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u2212<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 2.3014em;vertical-align: -0.93em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0359em\">\u03b7<\/span><span class=\"mord\"><span class=\"mopen nulldelimiter\"><\/span><span class=\"mfrac\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.3714em\"><span class=\"\" style=\"top: -2.2528em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"mord\"><span class=\"mord sqrt\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.8572em\"><span class=\"svg-align\" style=\"top: -3em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"mord\" style=\"padding-left: 0.833em\"><span class=\"mord\"><span class=\"mord accent\"><span class=\"vlist-t\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.6944em\"><span class=\"\" style=\"top: -3em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0359em\">v<\/span><\/span><span class=\"\" style=\"top: -3em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"accent-body\" style=\"left: -0.2222em\"><span class=\"mord\">^<\/span><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.2806em\"><span class=\"\" style=\"top: -2.55em;margin-left: -0.0359em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mathnormal mtight\">t<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"\" style=\"top: -2.8172em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"hide-tail\" style=\"min-width: 0.853em;height: 1.08em\"><\/p>\n<p>                     <\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.1828em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">&#043;<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mord mathnormal\">\u03f5<\/span><\/span><\/span><span class=\"\" style=\"top: -3.23em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"frac-line\" style=\"border-bottom-width: 0.04em\"><\/span><\/span><span class=\"\" style=\"top: -3.677em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"mord\"><span class=\"mord\"><span class=\"mord accent\"><span class=\"vlist-t\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.6944em\"><span class=\"\" style=\"top: -3em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"mord mathnormal\">m<\/span><\/span><span class=\"\" style=\"top: -3em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"accent-body\" style=\"left: -0.25em\"><span class=\"mord\">^<\/span><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.2806em\"><span class=\"\" style=\"top: -2.55em;margin-left: 0em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mathnormal mtight\">t<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.93em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><span class=\"mclose nulldelimiter\"><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/p>\n<\/li>\n<p>\u5176\u4e2d&#xff1a;<\/p>\n<ul>\n<li><span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">\n<p>          \u03b2<\/p>\n<p>          1<\/p>\n<p>         &#061;<\/p>\n<p>         0.9<\/p>\n<p>        \\\\beta_1&#061;0.9<\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.8889em;vertical-align: -0.1944em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.0528em\">\u03b2<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.3011em\"><span class=\"\" style=\"top: -2.55em;margin-left: -0.0528em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\">1<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.6444em\"><\/span><span class=\"mord\">0.9<\/span><\/span><\/span><\/span><\/span>, <span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\"><\/p>\n<p>          \u03b2<\/p>\n<p>          2<\/p>\n<p>         &#061;<\/p>\n<p>         0.999<\/p>\n<p>        \\\\beta_2&#061;0.999<\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.8889em;vertical-align: -0.1944em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.0528em\">\u03b2<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.3011em\"><span class=\"\" style=\"top: -2.55em;margin-left: -0.0528em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\">2<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.6444em\"><\/span><span class=\"mord\">0.999<\/span><\/span><\/span><\/span><\/span>, <span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\"><\/p>\n<p>         \u03f5<\/p>\n<p>         &#061;<\/p>\n<p>         1<\/p>\n<p>         e<\/p>\n<p>         \u2212<\/p>\n<p>         8<\/p>\n<p>        \\\\epsilon&#061;1e-8<\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.4306em\"><\/span><span class=\"mord mathnormal\">\u03f5<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.7278em;vertical-align: -0.0833em\"><\/span><span class=\"mord\">1<\/span><span class=\"mord mathnormal\">e<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u2212<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.6444em\"><\/span><span class=\"mord\">8<\/span><\/span><\/span><\/span><\/span>&#xff08;PyTorch\u9ed8\u8ba4\u503c&#xff09;<\/li>\n<li><span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\">\n<p>         \u03b7<\/p>\n<p>         &#061;<\/p>\n<p>         1<\/p>\n<p>         e<\/p>\n<p>         \u2212<\/p>\n<p>         3<\/p>\n<p>        \\\\eta&#061;1e-3<\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.625em;vertical-align: -0.1944em\"><\/span><span class=\"mord mathnormal\" style=\"margin-right: 0.0359em\">\u03b7<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.7278em;vertical-align: -0.0833em\"><\/span><span class=\"mord\">1<\/span><span class=\"mord mathnormal\">e<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><span class=\"mbin\">\u2212<\/span><span class=\"mspace\" style=\"margin-right: 0.2222em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.6444em\"><\/span><span class=\"mord\">3<\/span><\/span><\/span><\/span><\/span>&#xff08;\u5b66\u4e60\u7387&#xff0c;\u672c\u9879\u76ee\u9009\u62e9&#xff09;<\/li>\n<\/ul>\n<h6>\u4ee3\u7801\u8bbe\u8ba1\u601d\u8def&#xff1a;<\/h6>\n<ul>\n<li>\u4ea4\u53c9\u71b5\u635f\u5931&#xff1a;\u9002\u5408\u5206\u7c7b\u4efb\u52a1&#xff0c;\u5bf9\u9519\u8bef\u5206\u7c7b\u7684\u60e9\u7f5a\u66f4\u91cd<\/li>\n<li>Adam\u4f18\u5316\u5668&#xff1a;\u81ea\u9002\u5e94\u5b66\u4e60\u7387&#xff0c;\u8bad\u7ec3\u7a33\u5b9a\u6027\u597d&#xff0c;\u6536\u655b\u901f\u5ea6\u5feb<\/li>\n<li>\u5b66\u4e60\u73871e-3&#xff1a;\u5e38\u7528\u7684\u521d\u59cb\u5b66\u4e60\u7387&#xff0c;\u5e73\u8861\u4e86\u6536\u655b\u901f\u5ea6\u548c\u7a33\u5b9a\u6027<\/li>\n<\/ul>\n<h5>4.3.6 Dropout&#xff1a;\u9632\u6b62\u8fc7\u62df\u5408\u7684\u6b63\u5219\u5316\u6280\u672f<\/h5>\n<p>nn<span class=\"token punctuation\">.<\/span>Dropout<span class=\"token punctuation\">(<\/span><span class=\"token number\">0.2<\/span><span class=\"token punctuation\">)<\/span>  <span class=\"token comment\"># Dropout\u7387\u4e3a0.2<\/span><\/p>\n<h6>\u6570\u5b66\u539f\u7406\u4e0e\u63a8\u5bfc&#xff1a;<\/h6>\n<p>Dropout\u662f\u4e00\u79cd\u6b63\u5219\u5316\u6280\u672f&#xff0c;\u7528\u4e8e\u9632\u6b62\u8fc7\u62df\u5408&#xff0c;\u5728\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u968f\u673a\u5c06\u90e8\u5206\u795e\u7ecf\u5143\u7684\u8f93\u51fa\u7f6e\u4e3a0&#xff1a;<\/p>\n<p><span class=\"katex--display\"><span class=\"katex-display\"><span class=\"katex\"><span class=\"katex-mathml\"><\/p>\n<p>          y<\/p>\n<p>          i<\/p>\n<p>         &#061;<\/p>\n<p>          {<\/p>\n<p>               0<\/p>\n<p>               ,<\/p>\n<p>               with\u00a0probability\u00a0<\/p>\n<p>               p<\/p>\n<p>                 x<\/p>\n<p>                 i<\/p>\n<p>                 1<\/p>\n<p>                 \u2212<\/p>\n<p>                 p<\/p>\n<p>               ,<\/p>\n<p>              otherwise<\/p>\n<p>         y_i &#061; \\\\begin{cases} 0, &amp; \\\\text{with probability } p \\\\\\\\ \\\\frac{x_i}{1-p}, &amp; \\\\text{otherwise} \\\\end{cases} <\/p>\n<p>     <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.625em;vertical-align: -0.1944em\"><\/span><span class=\"mord\"><span class=\"mord mathnormal\" style=\"margin-right: 0.0359em\">y<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.3117em\"><span class=\"\" style=\"top: -2.55em;margin-left: -0.0359em;margin-right: 0.05em\"><span class=\"pstrut\" style=\"height: 2.7em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mathnormal mtight\">i<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.15em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 3em;vertical-align: -1.25em\"><\/span><span class=\"minner\"><span class=\"mopen delimcenter\" style=\"top: 0em\"><span class=\"delimsizing size4\">{<\/span><\/span><span class=\"mord\"><span class=\"mtable\"><span class=\"col-align-l\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.7146em\"><span class=\"\" style=\"top: -3.7146em\"><span class=\"pstrut\" style=\"height: 3.008em\"><\/span><span class=\"mord\"><span class=\"mord\">0<\/span><span class=\"mpunct\">,<\/span><\/span><\/span><span class=\"\" style=\"top: -2.2746em\"><span class=\"pstrut\" style=\"height: 3.008em\"><\/span><span class=\"mord\"><span class=\"mord\"><span class=\"mopen nulldelimiter\"><\/span><span class=\"mfrac\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.7115em\"><span class=\"\" style=\"top: -2.655em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mtight\">1<\/span><span class=\"mbin mtight\">\u2212<\/span><span class=\"mord mathnormal mtight\">p<\/span><\/span><\/span><\/span><span class=\"\" style=\"top: -3.23em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"frac-line\" style=\"border-bottom-width: 0.04em\"><\/span><\/span><span class=\"\" style=\"top: -3.4101em\"><span class=\"pstrut\" style=\"height: 3em\"><\/span><span class=\"sizing reset-size6 size3 mtight\"><span class=\"mord mtight\"><span class=\"mord mtight\"><span class=\"mord mathnormal mtight\">x<\/span><span class=\"msupsub\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.3281em\"><span class=\"\" style=\"top: -2.357em;margin-left: 0em;margin-right: 0.0714em\"><span class=\"pstrut\" style=\"height: 2.5em\"><\/span><span class=\"sizing reset-size3 size1 mtight\"><span class=\"mord mathnormal mtight\">i<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.143em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 0.4811em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><span class=\"mclose nulldelimiter\"><\/span><\/span><span class=\"mpunct\">,<\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.2146em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><span class=\"arraycolsep\" style=\"width: 1em\"><\/span><span class=\"col-align-l\"><span class=\"vlist-t vlist-t2\"><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.7146em\"><span class=\"\" style=\"top: -3.7146em\"><span class=\"pstrut\" style=\"height: 3.008em\"><\/span><span class=\"mord\"><span class=\"mord text\"><span class=\"mord\">with\u00a0probability\u00a0<\/span><\/span><span class=\"mord mathnormal\">p<\/span><\/span><\/span><span class=\"\" style=\"top: -2.2746em\"><span class=\"pstrut\" style=\"height: 3.008em\"><\/span><span class=\"mord\"><span class=\"mord text\"><span class=\"mord\">otherwise<\/span><\/span><\/span><\/span><\/span><span class=\"vlist-s\">\u200b<\/span><\/span><span class=\"vlist-r\"><span class=\"vlist\" style=\"height: 1.2146em\"><span class=\"\"><\/span><\/span><\/span><\/span><\/span><\/span><\/span><span class=\"mclose nulldelimiter\"><\/span><\/span><\/span><\/span><\/span><\/span><\/span><\/p>\n<p>\u5176\u4e2d<span class=\"katex--inline\"><span class=\"katex\"><span class=\"katex-mathml\"><\/p>\n<p>        p<\/p>\n<p>        &#061;<\/p>\n<p>        0.2<\/p>\n<p>       p&#061;0.2<\/p>\n<p>    <\/span><span class=\"katex-html\"><span class=\"base\"><span class=\"strut\" style=\"height: 0.625em;vertical-align: -0.1944em\"><\/span><span class=\"mord mathnormal\">p<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><span class=\"mrel\">&#061;<\/span><span class=\"mspace\" style=\"margin-right: 0.2778em\"><\/span><\/span><span class=\"base\"><span class=\"strut\" style=\"height: 0.6444em\"><\/span><span class=\"mord\">0.2<\/span><\/span><\/span><\/span><\/span>\u662fdropout\u7387\u3002<\/p>\n<h6>\u4ee3\u7801\u8bbe\u8ba1\u601d\u8def&#xff1a;<\/h6>\n<ul>\n<li>dropout\u73870.2&#xff1a;\u8f83\u5c0f\u7684dropout\u7387&#xff0c;\u5728\u9632\u6b62\u8fc7\u62df\u5408\u7684\u540c\u65f6\u4fdd\u7559\u5927\u90e8\u5206\u4fe1\u606f<\/li>\n<li>\u5168\u8fde\u63a5\u5c42\u540e\u4f7f\u7528&#xff1a;\u5168\u8fde\u63a5\u5c42\u53c2\u6570\u591a&#xff0c;\u5bb9\u6613\u8fc7\u62df\u5408&#xff0c;\u662fdropout\u7684\u4e3b\u8981\u5e94\u7528\u4f4d\u7f6e<\/li>\n<li>\u6d4b\u8bd5\u65f6\u4e0d\u4f7f\u7528&#xff1a;PyTorch\u7684model.eval()\u4f1a\u81ea\u52a8\u5173\u95eddropout<\/li>\n<\/ul>\n<h6>\u4e0e\u9879\u76ee\u7684\u7ed3\u5408&#xff1a;<\/h6>\n<p>\u5728\u672c\u9879\u76ee\u4e2d&#xff0c;dropout\u5c42\u5e94\u7528\u4e8e&#xff1a;<\/p>\n<ul>\n<li>512\u2192256\u5168\u8fde\u63a5\u5c42\u4e4b\u540e<\/li>\n<li>256\u2192128\u5168\u8fde\u63a5\u5c42\u4e4b\u540e<\/li>\n<li>128\u219210\u5168\u8fde\u63a5\u5c42\u4e4b\u540e&#xff08;\u8f93\u51fa\u5c42\u4e4b\u524d&#xff09;<\/li>\n<\/ul>\n<h3>\u4e94\u3001\u9879\u76ee\u4f7f\u7528\u6307\u5357<\/h3>\n<h4>5.1 \u73af\u5883\u642d\u5efa<\/h4>\n<li>\n<p>\u5b89\u88c5\u5fc5\u8981\u7684\u4f9d\u8d56&#xff1a;<\/p>\n<p> pip <span class=\"token function\">install<\/span> torch torchvision matplotlib seaborn scikit-learn tqdm\n <\/li>\n<li>\n<p>\u786e\u4fdd\u6709\u8db3\u591f\u7684\u8ba1\u7b97\u8d44\u6e90&#xff08;GPU\u63a8\u8350&#xff09;<\/p>\n<\/li>\n<h4>5.2 \u6570\u636e\u51c6\u5907<\/h4>\n<li>CIFAR10\u6570\u636e\u96c6\u4f1a\u81ea\u52a8\u4e0b\u8f7d\u5230\u9879\u76ee\u7684.\/data\u76ee\u5f55<\/li>\n<li>\u9996\u6b21\u8fd0\u884c\u65f6&#xff0c;\u5c06\u4ee3\u7801\u4e2d\u7684download&#061;False\u6539\u4e3adownload&#061;True<\/li>\n<h4>5.3 \u8fd0\u884c\u4ee3\u7801<\/h4>\n<li>\n<p>\u76f4\u63a5\u8fd0\u884c\u4e3b\u7a0b\u5e8f&#xff1a;<\/p>\n<p> python CNN_DNN.py\n <\/li>\n<li>\n<p>\u4ee3\u7801\u6267\u884c\u6d41\u7a0b&#xff1a;<\/p>\n<ul>\n<li>\u52a0\u8f7d\u5e76\u9884\u5904\u7406\u6570\u636e<\/li>\n<li>\u521d\u59cb\u5316\u6a21\u578b<\/li>\n<li>\u8bad\u7ec3\u6a21\u578b&#xff08;10\u8f6e&#xff09;<\/li>\n<li>\u7ed8\u5236\u8bad\u7ec3\u635f\u5931\u66f2\u7ebf<\/li>\n<li>\u6d4b\u8bd5\u6a21\u578b<\/li>\n<li>\u751f\u6210\u6df7\u6dc6\u77e9\u9635<\/li>\n<li>\u5c55\u793a\u9519\u8bef\u5206\u7c7b\u548c\u6b63\u786e\u5206\u7c7b\u7684\u6837\u672c<\/li>\n<li>\u5c55\u793a\u9884\u6d4b\u6982\u7387\u5206\u5e03<\/li>\n<\/ul>\n<\/li>\n<h4>5.4 \u7ed3\u679c\u67e5\u770b<\/h4>\n<li>\n<p>\u7ec8\u7aef\u8f93\u51fa&#xff1a;<\/p>\n<ul>\n<li>\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u7684\u635f\u5931\u4fe1\u606f<\/li>\n<li>\u6d4b\u8bd5\u51c6\u786e\u7387<\/li>\n<\/ul>\n<\/li>\n<li>\n<p>\u751f\u6210\u7684\u53ef\u89c6\u5316\u6587\u4ef6&#xff1a;<\/p>\n<ul>\n<li>training_loss_curve.png&#xff1a;\u8bad\u7ec3\u635f\u5931\u66f2\u7ebf<\/li>\n<li>confusion_matrix.png&#xff1a;\u6df7\u6dc6\u77e9\u9635<\/li>\n<li>misclassified_samples.png&#xff1a;\u9519\u8bef\u5206\u7c7b\u6837\u672c<\/li>\n<li>correctly_classified_samples.png&#xff1a;\u6b63\u786e\u5206\u7c7b\u6837\u672c<\/li>\n<li>prediction_probabilities.png&#xff1a;\u9884\u6d4b\u6982\u7387\u5206\u5e03<\/li>\n<\/ul>\n<\/li>\n<h4>5.5 \u53c2\u6570\u8c03\u6574<\/h4>\n<li>\u8bad\u7ec3\u8f6e\u6570&#xff1a;\u4fee\u6539epochs\u53c2\u6570<\/li>\n<li>\u5b66\u4e60\u7387&#xff1a;\u4fee\u6539optimizer\u7684lr\u53c2\u6570<\/li>\n<li>\u6279\u91cf\u5927\u5c0f&#xff1a;\u4fee\u6539DataLoader\u7684batch_size\u53c2\u6570<\/li>\n<li>\u6a21\u578b\u7ed3\u6784&#xff1a;\u4fee\u6539CNN_DNN_Model\u7c7b\u4e2d\u7684\u7f51\u7edc\u5c42\u914d\u7f6e<\/li>\n<h3>6. \u8fd0\u884c\u6548\u679c<\/h3>\n<p>&#x1f680; \u5f00\u59cb\u8bad\u7ec3\u6a21\u578b&#8230;<\/p>\n<p>\u2705 Epoch [1\/50] \u5b8c\u6210 | \u672c\u8f6e\u5e73\u5747\u635f\u5931: 1.7453 | \u8bad\u7ec3\u51c6\u786e\u7387: 47.41% | \u9a8c\u8bc1\u51c6\u786e\u7387: 48.15%<\/p>\n<p>&#x1f4be; \u9a8c\u8bc1\u51c6\u786e\u7387\u63d0\u5347\u81f3 48.15%&#xff0c;\u5df2\u4fdd\u5b58\u6700\u4f73\u6a21\u578b\u5230 best_model.pth<\/p>\n<p> (  \u6b64  \u5904  \u7701  \u7565  \u4e00  \u90e8  \u5206  )<\/p>\n<p>\u2705 Epoch [35\/50] \u5b8c\u6210 | \u672c\u8f6e\u5e73\u5747\u635f\u5931: 0.5103 | \u8bad\u7ec3\u51c6\u786e\u7387: 84.67% | \u9a8c\u8bc1\u51c6\u786e\u7387: 81.49%<\/p>\n<p>\u2705 Epoch [36\/50] \u5b8c\u6210 | \u672c\u8f6e\u5e73\u5747\u635f\u5931: 0.5094 | \u8bad\u7ec3\u51c6\u786e\u7387: 84.18% | \u9a8c\u8bc1\u51c6\u786e\u7387: 81.59%<\/p>\n<p>&#x1f4be; \u9a8c\u8bc1\u51c6\u786e\u7387\u63d0\u5347\u81f3 81.59%&#xff0c;\u5df2\u4fdd\u5b58\u6700\u4f73\u6a21\u578b\u5230 best_model.pth<\/p>\n<p>\u2705 Epoch [37\/50] \u5b8c\u6210 | \u672c\u8f6e\u5e73\u5747\u635f\u5931: 0.5086 | \u8bad\u7ec3\u51c6\u786e\u7387: 85.60% | \u9a8c\u8bc1\u51c6\u786e\u7387: 81.82%<\/p>\n<p>&#x1f4be; \u9a8c\u8bc1\u51c6\u786e\u7387\u63d0\u5347\u81f3 81.82%&#xff0c;\u5df2\u4fdd\u5b58\u6700\u4f73\u6a21\u578b\u5230 best_model.pth<\/p>\n<p>\u2705 Epoch [38\/50] \u5b8c\u6210 | \u672c\u8f6e\u5e73\u5747\u635f\u5931: 0.5008 | \u8bad\u7ec3\u51c6\u786e\u7387: 86.28% | \u9a8c\u8bc1\u51c6\u786e\u7387: 82.10%<\/p>\n<p>&#x1f4be; \u9a8c\u8bc1\u51c6\u786e\u7387\u63d0\u5347\u81f3 82.10%&#xff0c;\u5df2\u4fdd\u5b58\u6700\u4f73\u6a21\u578b\u5230 best_model.pth<\/p>\n<p>\u2705 Epoch [39\/50] \u5b8c\u6210 | \u672c\u8f6e\u5e73\u5747\u635f\u5931: 0.4955 | \u8bad\u7ec3\u51c6\u786e\u7387: 84.67% | \u9a8c\u8bc1\u51c6\u786e\u7387: 81.71%<\/p>\n<p>\u2705 Epoch [40\/50] \u5b8c\u6210 | \u672c\u8f6e\u5e73\u5747\u635f\u5931: 0.5008 | \u8bad\u7ec3\u51c6\u786e\u7387: 85.01% | \u9a8c\u8bc1\u51c6\u786e\u7387: 82.41%<\/p>\n<p>&#x1f4be; \u9a8c\u8bc1\u51c6\u786e\u7387\u63d0\u5347\u81f3 82.41%&#xff0c;\u5df2\u4fdd\u5b58\u6700\u4f73\u6a21\u578b\u5230 best_model.pth<\/p>\n<p>\u2705 Epoch [41\/50] \u5b8c\u6210 | \u672c\u8f6e\u5e73\u5747\u635f\u5931: 0.4960 | \u8bad\u7ec3\u51c6\u786e\u7387: 84.86% | \u9a8c\u8bc1\u51c6\u786e\u7387: 81.71%<\/p>\n<p>\u2705 Epoch [42\/50] \u5b8c\u6210 | \u672c\u8f6e\u5e73\u5747\u635f\u5931: 0.4955 | \u8bad\u7ec3\u51c6\u786e\u7387: 86.08% | \u9a8c\u8bc1\u51c6\u786e\u7387: 81.99%<\/p>\n<p>\u2705 Epoch [43\/50] \u5b8c\u6210 | \u672c\u8f6e\u5e73\u5747\u635f\u5931: 0.4778 | \u8bad\u7ec3\u51c6\u786e\u7387: 85.60% | \u9a8c\u8bc1\u51c6\u786e\u7387: 80.98%<\/p>\n<p>\u2705 Epoch [44\/50] \u5b8c\u6210 | \u672c\u8f6e\u5e73\u5747\u635f\u5931: 0.4862 | \u8bad\u7ec3\u51c6\u786e\u7387: 85.60% | \u9a8c\u8bc1\u51c6\u786e\u7387: 82.59%<\/p>\n<p>&#x1f4be; \u9a8c\u8bc1\u51c6\u786e\u7387\u63d0\u5347\u81f3 82.59%&#xff0c;\u5df2\u4fdd\u5b58\u6700\u4f73\u6a21\u578b\u5230 best_model.pth<\/p>\n<p>\u2705 Epoch [45\/50] \u5b8c\u6210 | \u672c\u8f6e\u5e73\u5747\u635f\u5931: 0.4764 | \u8bad\u7ec3\u51c6\u786e\u7387: 85.45% | \u9a8c\u8bc1\u51c6\u786e\u7387: 82.34%<\/p>\n<p>\u2705 Epoch [46\/50] \u5b8c\u6210 | \u672c\u8f6e\u5e73\u5747\u635f\u5931: 0.4776 | \u8bad\u7ec3\u51c6\u786e\u7387: 84.77% | \u9a8c\u8bc1\u51c6\u786e\u7387: 81.36%<\/p>\n<p>\u2705 Epoch [47\/50] \u5b8c\u6210 | \u672c\u8f6e\u5e73\u5747\u635f\u5931: 0.4796 | \u8bad\u7ec3\u51c6\u786e\u7387: 84.18% | \u9a8c\u8bc1\u51c6\u786e\u7387: 81.78%<\/p>\n<p>\u2705 Epoch [48\/50] \u5b8c\u6210 | \u672c\u8f6e\u5e73\u5747\u635f\u5931: 0.4672 | \u8bad\u7ec3\u51c6\u786e\u7387: 86.23% | \u9a8c\u8bc1\u51c6\u786e\u7387: 82.07%<\/p>\n<p>\u2705 Epoch [49\/50] \u5b8c\u6210 | \u672c\u8f6e\u5e73\u5747\u635f\u5931: 0.4719 | \u8bad\u7ec3\u51c6\u786e\u7387: 85.84% | \u9a8c\u8bc1\u51c6\u786e\u7387: 82.10%<\/p>\n<p>\u2705 Epoch [50\/50] \u5b8c\u6210 | \u672c\u8f6e\u5e73\u5747\u635f\u5931: 0.4669 | \u8bad\u7ec3\u51c6\u786e\u7387: 86.28% | \u9a8c\u8bc1\u51c6\u786e\u7387: 82.79%<\/p>\n<p>&#x1f4be; \u9a8c\u8bc1\u51c6\u786e\u7387\u63d0\u5347\u81f3 82.79%&#xff0c;\u5df2\u4fdd\u5b58\u6700\u4f73\u6a21\u578b\u5230 best_model.pth<\/p>\n<p>&#x1f389; \u6240\u6709\u8bad\u7ec3\u8f6e\u6570\u5b8c\u6210&#xff01;<br \/>\n&#x1f3c6; \u6700\u4f73\u9a8c\u8bc1\u51c6\u786e\u7387: 82.79%<br \/>\n&#x1f4c1; \u6700\u4f73\u6a21\u578b\u5df2\u4fdd\u5b58\u5230: best_model.pth<\/p>\n<p>&#x1f4c8; \u7ed8\u5236\u8bad\u7ec3\u635f\u5931\u66f2\u7ebf&#8230;<\/p>\n<p>\u2705 \u8bad\u7ec3\u635f\u5931\u66f2\u7ebf\u5df2\u4fdd\u5b58\u5e76\u663e\u793a<br \/>\n&#x1f4ca; \u7ed8\u5236\u51c6\u786e\u7387\u5bf9\u6bd4\u66f2\u7ebf&#8230;<\/p>\n<p>\u2705 \u8bad\u7ec3\u548c\u9a8c\u8bc1\u51c6\u786e\u7387\u5bf9\u6bd4\u5df2\u4fdd\u5b58\u5e76\u663e\u793a<\/p>\n<p>&#x1f9ea; \u6d4b\u8bd5\u6a21\u578b\u6027\u80fd&#8230;<\/p>\n<p>&#x1f4ca; \u6d4b\u8bd5\u96c6\u51c6\u786e\u7387: 82.29%<\/p>\n<p>&#x1f50d; \u751f\u6210\u6df7\u6dc6\u77e9\u9635&#8230;<\/p>\n<p>\u2705 \u6df7\u6dc6\u77e9\u9635\u5df2\u4fdd\u5b58\u5e76\u663e\u793a<\/p>\n<p>\u274c \u5c55\u793a\u9519\u8bef\u5206\u7c7b\u6837\u672c&#8230;<\/p>\n<p>\u2705 \u9519\u8bef\u5206\u7c7b\u6837\u672c\u5df2\u4fdd\u5b58\u5e76\u663e\u793a<\/p>\n<p>\u2705 \u5c55\u793a\u6b63\u786e\u5206\u7c7b\u6837\u672c&#8230;<\/p>\n<p>\u2705 \u6b63\u786e\u5206\u7c7b\u6837\u672c\u5df2\u4fdd\u5b58\u5e76\u663e\u793a<\/p>\n<p>&#x1f4ca; \u5c55\u793a\u9884\u6d4b\u6982\u7387\u5206\u5e03&#8230;<\/p>\n<p>\u2705 \u9884\u6d4b\u6982\u7387\u5206\u5e03\u5df2\u4fdd\u5b58\u5e76\u663e\u793a<\/p>\n<h3>7. \u9879\u76ee\u6e90\u7801<\/h3>\n<p>https:\/\/download.csdn.net\/download\/qq_68716132\/92561378?spm&#061;1001.2014.3001.5501<\/p>\n<h3>8. \u603b\u7ed3<\/h3>\n<ul>\n<li>\u603b\u8bad\u7ec3\u8f6e\u6570: 50<\/li>\n<li>\u6700\u7ec8\u8bad\u7ec3\u635f\u5931: 0.4669<\/li>\n<li>\u6700\u7ec8\u8bad\u7ec3\u51c6\u786e\u7387: 86.28%<\/li>\n<li>\u6700\u7ec8\u9a8c\u8bc1\u51c6\u786e\u7387: 82.79%<\/li>\n<li>\u6700\u4f73\u9a8c\u8bc1\u51c6\u786e\u7387: 82.79%<\/li>\n<\/ul>\n<p>\u672c\u9879\u76ee\u7ed3\u5408\u4e86CNN\u548cDNN\u7684\u4f18\u52bf&#xff0c;\u5229\u7528CNN\u63d0\u53d6\u56fe\u50cf\u7684\u5c40\u90e8\u7279\u5f81&#xff0c;\u518d\u901a\u8fc7DNN\u8fdb\u884c\u5206\u7c7b\u3002\u8fd9\u79cd\u67b6\u6784\u5728\u56fe\u50cf\u5206\u7c7b\u4efb\u52a1\u4e2d\u8868\u73b0\u51fa\u8272&#xff0c;\u7279\u522b\u662f\u5bf9\u4e8eCIFAR10\u8fd9\u6837\u7684\u5c0f\u578b\u56fe\u50cf\u6570\u636e\u96c6\u3002\u901a\u8fc7\u8be6\u7ec6\u7684\u6ce8\u91ca\u548c\u53ef\u89c6\u5316\u529f\u80fd&#xff0c;\u9879\u76ee\u4ee3\u7801\u5177\u6709\u826f\u597d\u7684\u53ef\u8bfb\u6027\u548c\u53ef\u7ef4\u62a4\u6027&#xff0c;\u9002\u5408\u4f5c\u4e3a\u6df1\u5ea6\u5b66\u4e60\u521d\u5b66\u8005\u7684\u5b66\u4e60\u793a\u4f8b\u3002<\/p>\n","protected":false},"excerpt":{"rendered":"<p>\u6df1\u5ea6\u5b66\u4e60\u7b14\u8bb0&#xff1a;\u4eceANN\u5230CNNDNN\u7684\u56fe\u50cf\u5206\u7c7b\u5b9e\u8df5<br \/>\n\u4e00\u3001\u6df1\u5ea6\u5b66\u4e60\u57fa\u7840&#xff1a;ANN\u3001CNN\u4e0eDNN\u7684\u6838\u5fc3\u539f\u7406<br \/>\n1.1 \u4eba\u5de5\u795e\u7ecf\u7f51\u7edc&#xff08;ANN&#xff09;&#xff1a;\u6df1\u5ea6\u5b66\u4e60\u7684\u8d77\u70b9<br \/>\n1.1.1 \u4ec0\u4e48\u662f\u4eba\u5de5\u795e\u7ecf\u7f51\u7edc&#xff1f;<br \/>\n\u4eba\u5de5\u795e\u7ecf\u7f51\u7edc&#xff08;Artificial Neural Network, ANN&#xff09;\u662f\u4e00\u79cd\u6a21\u4eff\u751f\u7269\u795e\u7ecf\u7f51\u7edc\u7ed3\u6784\u548c\u529f\u80fd\u7684\u6570\u5b66\u6a21\u578b&#xff0c;\u7531\u5927\u91cf\u7684\u795e\u7ecf\u5143\u76f8\u4e92\u8fde\u63a5\u800c\u6210\u3002<br \/>\n1.1.<\/p>\n","protected":false},"author":2,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[1],"tags":[6342,2394,86,305],"topic":[],"class_list":["post-60163","post","type-post","status-publish","format-standard","hentry","category-server","tag-dnn","tag-cnn","tag-86","tag-305"],"yoast_head":"<!-- This site is optimized with the Yoast SEO plugin v20.3 - https:\/\/yoast.com\/wordpress\/plugins\/seo\/ -->\n<title>\u8d85\u8be6\u7ec6\u6df1\u5ea6\u5b66\u4e60\u7b14\u8bb0\uff1a\u4eceANN\u5230CNN+DNN\u7684\u56fe\u50cf\u5206\u7c7b\u5b9e\u8df5\uff0c\u5168\u8bb2\u660e\u767d\u5566\uff01 - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3<\/title>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/www.wsisp.com\/helps\/60163.html\" \/>\n<meta property=\"og:locale\" content=\"zh_CN\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"\u8d85\u8be6\u7ec6\u6df1\u5ea6\u5b66\u4e60\u7b14\u8bb0\uff1a\u4eceANN\u5230CNN+DNN\u7684\u56fe\u50cf\u5206\u7c7b\u5b9e\u8df5\uff0c\u5168\u8bb2\u660e\u767d\u5566\uff01 - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3\" \/>\n<meta property=\"og:description\" content=\"\u6df1\u5ea6\u5b66\u4e60\u7b14\u8bb0&#xff1a;\u4eceANN\u5230CNNDNN\u7684\u56fe\u50cf\u5206\u7c7b\u5b9e\u8df5 \u4e00\u3001\u6df1\u5ea6\u5b66\u4e60\u57fa\u7840&#xff1a;ANN\u3001CNN\u4e0eDNN\u7684\u6838\u5fc3\u539f\u7406 1.1 \u4eba\u5de5\u795e\u7ecf\u7f51\u7edc&#xff08;ANN&#xff09;&#xff1a;\u6df1\u5ea6\u5b66\u4e60\u7684\u8d77\u70b9 1.1.1 \u4ec0\u4e48\u662f\u4eba\u5de5\u795e\u7ecf\u7f51\u7edc&#xff1f; \u4eba\u5de5\u795e\u7ecf\u7f51\u7edc&#xff08;Artificial Neural Network, ANN&#xff09;\u662f\u4e00\u79cd\u6a21\u4eff\u751f\u7269\u795e\u7ecf\u7f51\u7edc\u7ed3\u6784\u548c\u529f\u80fd\u7684\u6570\u5b66\u6a21\u578b&#xff0c;\u7531\u5927\u91cf\u7684\u795e\u7ecf\u5143\u76f8\u4e92\u8fde\u63a5\u800c\u6210\u3002 1.1.\" \/>\n<meta property=\"og:url\" content=\"https:\/\/www.wsisp.com\/helps\/60163.html\" \/>\n<meta property=\"og:site_name\" content=\"\u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3\" \/>\n<meta property=\"article:published_time\" content=\"2026-01-14T16:57:17+00:00\" \/>\n<meta name=\"author\" content=\"admin\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"\u4f5c\u8005\" \/>\n\t<meta name=\"twitter:data1\" content=\"admin\" \/>\n\t<meta name=\"twitter:label2\" content=\"\u9884\u8ba1\u9605\u8bfb\u65f6\u95f4\" \/>\n\t<meta name=\"twitter:data2\" content=\"12 \u5206\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\/\/schema.org\",\"@graph\":[{\"@type\":\"WebPage\",\"@id\":\"https:\/\/www.wsisp.com\/helps\/60163.html\",\"url\":\"https:\/\/www.wsisp.com\/helps\/60163.html\",\"name\":\"\u8d85\u8be6\u7ec6\u6df1\u5ea6\u5b66\u4e60\u7b14\u8bb0\uff1a\u4eceANN\u5230CNN+DNN\u7684\u56fe\u50cf\u5206\u7c7b\u5b9e\u8df5\uff0c\u5168\u8bb2\u660e\u767d\u5566\uff01 - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3\",\"isPartOf\":{\"@id\":\"https:\/\/www.wsisp.com\/helps\/#website\"},\"datePublished\":\"2026-01-14T16:57:17+00:00\",\"dateModified\":\"2026-01-14T16:57:17+00:00\",\"author\":{\"@id\":\"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/358e386c577a3ab51c4493330a20ad41\"},\"breadcrumb\":{\"@id\":\"https:\/\/www.wsisp.com\/helps\/60163.html#breadcrumb\"},\"inLanguage\":\"zh-Hans\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\/\/www.wsisp.com\/helps\/60163.html\"]}]},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\/\/www.wsisp.com\/helps\/60163.html#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"\u9996\u9875\",\"item\":\"https:\/\/www.wsisp.com\/helps\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"\u8d85\u8be6\u7ec6\u6df1\u5ea6\u5b66\u4e60\u7b14\u8bb0\uff1a\u4eceANN\u5230CNN+DNN\u7684\u56fe\u50cf\u5206\u7c7b\u5b9e\u8df5\uff0c\u5168\u8bb2\u660e\u767d\u5566\uff01\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\/\/www.wsisp.com\/helps\/#website\",\"url\":\"https:\/\/www.wsisp.com\/helps\/\",\"name\":\"\u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3\",\"description\":\"\u9999\u6e2f\u670d\u52a1\u5668_\u9999\u6e2f\u4e91\u670d\u52a1\u5668\u8d44\u8baf_\u670d\u52a1\u5668\u5e2e\u52a9\u6587\u6863_\u670d\u52a1\u5668\u6559\u7a0b\",\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\/\/www.wsisp.com\/helps\/?s={search_term_string}\"},\"query-input\":\"required name=search_term_string\"}],\"inLanguage\":\"zh-Hans\"},{\"@type\":\"Person\",\"@id\":\"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/358e386c577a3ab51c4493330a20ad41\",\"name\":\"admin\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"zh-Hans\",\"@id\":\"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/image\/\",\"url\":\"https:\/\/gravatar.wp-china-yes.net\/avatar\/?s=96&d=mystery\",\"contentUrl\":\"https:\/\/gravatar.wp-china-yes.net\/avatar\/?s=96&d=mystery\",\"caption\":\"admin\"},\"sameAs\":[\"http:\/\/wp.wsisp.com\"],\"url\":\"https:\/\/www.wsisp.com\/helps\/author\/admin\"}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"\u8d85\u8be6\u7ec6\u6df1\u5ea6\u5b66\u4e60\u7b14\u8bb0\uff1a\u4eceANN\u5230CNN+DNN\u7684\u56fe\u50cf\u5206\u7c7b\u5b9e\u8df5\uff0c\u5168\u8bb2\u660e\u767d\u5566\uff01 - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/www.wsisp.com\/helps\/60163.html","og_locale":"zh_CN","og_type":"article","og_title":"\u8d85\u8be6\u7ec6\u6df1\u5ea6\u5b66\u4e60\u7b14\u8bb0\uff1a\u4eceANN\u5230CNN+DNN\u7684\u56fe\u50cf\u5206\u7c7b\u5b9e\u8df5\uff0c\u5168\u8bb2\u660e\u767d\u5566\uff01 - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3","og_description":"\u6df1\u5ea6\u5b66\u4e60\u7b14\u8bb0&#xff1a;\u4eceANN\u5230CNNDNN\u7684\u56fe\u50cf\u5206\u7c7b\u5b9e\u8df5 \u4e00\u3001\u6df1\u5ea6\u5b66\u4e60\u57fa\u7840&#xff1a;ANN\u3001CNN\u4e0eDNN\u7684\u6838\u5fc3\u539f\u7406 1.1 \u4eba\u5de5\u795e\u7ecf\u7f51\u7edc&#xff08;ANN&#xff09;&#xff1a;\u6df1\u5ea6\u5b66\u4e60\u7684\u8d77\u70b9 1.1.1 \u4ec0\u4e48\u662f\u4eba\u5de5\u795e\u7ecf\u7f51\u7edc&#xff1f; \u4eba\u5de5\u795e\u7ecf\u7f51\u7edc&#xff08;Artificial Neural Network, ANN&#xff09;\u662f\u4e00\u79cd\u6a21\u4eff\u751f\u7269\u795e\u7ecf\u7f51\u7edc\u7ed3\u6784\u548c\u529f\u80fd\u7684\u6570\u5b66\u6a21\u578b&#xff0c;\u7531\u5927\u91cf\u7684\u795e\u7ecf\u5143\u76f8\u4e92\u8fde\u63a5\u800c\u6210\u3002 1.1.","og_url":"https:\/\/www.wsisp.com\/helps\/60163.html","og_site_name":"\u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3","article_published_time":"2026-01-14T16:57:17+00:00","author":"admin","twitter_card":"summary_large_image","twitter_misc":{"\u4f5c\u8005":"admin","\u9884\u8ba1\u9605\u8bfb\u65f6\u95f4":"12 \u5206"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"WebPage","@id":"https:\/\/www.wsisp.com\/helps\/60163.html","url":"https:\/\/www.wsisp.com\/helps\/60163.html","name":"\u8d85\u8be6\u7ec6\u6df1\u5ea6\u5b66\u4e60\u7b14\u8bb0\uff1a\u4eceANN\u5230CNN+DNN\u7684\u56fe\u50cf\u5206\u7c7b\u5b9e\u8df5\uff0c\u5168\u8bb2\u660e\u767d\u5566\uff01 - \u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3","isPartOf":{"@id":"https:\/\/www.wsisp.com\/helps\/#website"},"datePublished":"2026-01-14T16:57:17+00:00","dateModified":"2026-01-14T16:57:17+00:00","author":{"@id":"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/358e386c577a3ab51c4493330a20ad41"},"breadcrumb":{"@id":"https:\/\/www.wsisp.com\/helps\/60163.html#breadcrumb"},"inLanguage":"zh-Hans","potentialAction":[{"@type":"ReadAction","target":["https:\/\/www.wsisp.com\/helps\/60163.html"]}]},{"@type":"BreadcrumbList","@id":"https:\/\/www.wsisp.com\/helps\/60163.html#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"\u9996\u9875","item":"https:\/\/www.wsisp.com\/helps"},{"@type":"ListItem","position":2,"name":"\u8d85\u8be6\u7ec6\u6df1\u5ea6\u5b66\u4e60\u7b14\u8bb0\uff1a\u4eceANN\u5230CNN+DNN\u7684\u56fe\u50cf\u5206\u7c7b\u5b9e\u8df5\uff0c\u5168\u8bb2\u660e\u767d\u5566\uff01"}]},{"@type":"WebSite","@id":"https:\/\/www.wsisp.com\/helps\/#website","url":"https:\/\/www.wsisp.com\/helps\/","name":"\u7f51\u7855\u4e92\u8054\u5e2e\u52a9\u4e2d\u5fc3","description":"\u9999\u6e2f\u670d\u52a1\u5668_\u9999\u6e2f\u4e91\u670d\u52a1\u5668\u8d44\u8baf_\u670d\u52a1\u5668\u5e2e\u52a9\u6587\u6863_\u670d\u52a1\u5668\u6559\u7a0b","potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/www.wsisp.com\/helps\/?s={search_term_string}"},"query-input":"required name=search_term_string"}],"inLanguage":"zh-Hans"},{"@type":"Person","@id":"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/358e386c577a3ab51c4493330a20ad41","name":"admin","image":{"@type":"ImageObject","inLanguage":"zh-Hans","@id":"https:\/\/www.wsisp.com\/helps\/#\/schema\/person\/image\/","url":"https:\/\/gravatar.wp-china-yes.net\/avatar\/?s=96&d=mystery","contentUrl":"https:\/\/gravatar.wp-china-yes.net\/avatar\/?s=96&d=mystery","caption":"admin"},"sameAs":["http:\/\/wp.wsisp.com"],"url":"https:\/\/www.wsisp.com\/helps\/author\/admin"}]}},"_links":{"self":[{"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/posts\/60163","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/comments?post=60163"}],"version-history":[{"count":0,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/posts\/60163\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/media?parent=60163"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/categories?post=60163"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/tags?post=60163"},{"taxonomy":"topic","embeddable":true,"href":"https:\/\/www.wsisp.com\/helps\/wp-json\/wp\/v2\/topic?post=60163"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}