![\"\"](\"https:\/\/www.wsisp.com\/helps\/wp-content\/uploads\/2025\/05\/20250430220238-68129dfe5c79e.png\")
![\"\"](\"https:\/\/www.wsisp.com\/helps\/wp-content\/uploads\/2025\/05\/20250430220238-68129dfe6fde2.png\")
<\/p>\n\u6570\u636e\u96c6：\u516c\u5f00\u7684World Happiness Report | Kaggle\u4e2d\u7684happiness dataset2017.<\/p>\n
\u76ee\u6807：\u57fa\u4e8eGDP\u503c\u9884\u6d4b\u5e78\u798f\u6307\u6570\u3002（\u5355\u7279\u5f81\u9884\u6d4b）<\/p>\n
\u4ee3\u7801：<\/p>\n
\u6587\u4ef6\u4e00：prepare_for_traning.py<\/p>\n
"""\u7528\u4e8e\u79d1\u5b66\u8ba1\u7b97\u7684\u4e00\u4e2a\u5e93，\u63d0\u4f9b\u4e86\u591a\u7ef4\u6570\u7ec4\u5bf9\u8c61\u4ee5\u53ca\u64cd\u4f5c\u51fd\u6570"""
\nfrom utils.features import prepare_for_training
\n"""\u6570\u636e\u9884\u5904\u7406\u7684\u4e00\u4e2a\u79c1\u5e93"""<\/p>\n
class LinearRegression:
\n def __init__(self,data,labels,polynomial_degree = 0,sinusoid_degree = 0,normalize_data = True):
\n """
\n \u8fdb\u884c\u9884\u5904\u7406\u64cd\u4f5c
\n :param data:
\n :param labels:
\n :param polynomial_degree:
\n :param sinusoid_degree:
\n :param normalize_data:
\n """
\n (data_processed,
\n features_mean,
\n features_deviation) = prepare_for_training(data,polynomial_degree = 0,sinusoid_degree = 0,normalize_data = True)
\n self.data = data_processed
\n self.labels = labels
\n self.features_mean = features_mean
\n self.features_deviation = features_deviation
\n self.polynomial_degree = polynomial_degree
\n self.sinusoid_degree = sinusoid_degree
\n self.normalize_data = normalize_data
\n num_features = self.data.shape[1]
\n self.theta = np.zeros((num_features,1))
\n """"\u6570\u636e，\u5b66\u4e60\u7387，\u8bad\u7ec3\u6b21\u6570"""
\n def train(self,alpha,num_iterations = 500):
\n """\u8bad\u7ec3\u6a21\u5757：\u68af\u5ea6\u4e0b\u964d"""
\n cost_history = self.gradient_descent(alpha,num_iterations)
\n return self.theta,cost_history<\/p>\n
def gradient_descent(self,alpha,num_iterations):
\n """\u8fed\u4ee3\u6a21\u5757"""
\n cost_history = []
\n for _ in range(num_iterations):
\n self.gradient_step(alpha)
\n cost_history.append(self.cost_function(self.data,self.labels))
\n return cost_history<\/p>\n
def gradient_step(self,alpha):
\n """
\n \u68af\u5ea6\u4e0b\u964d\u53c2\u6570\u66f4\u65b0\u7b97\u6cd5，\u77e9\u9635\u8ba1\u7b97，\u4f7f\u7528\u5c0f\u6279\u91cf\u68af\u5ea6\u4e0b\u964d\u7b97\u6cd5
\n :param self:
\n :param alpha:
\n :return:
\n """
\n num_examples = self.data.shape[0]
\n prediction = LinearRegression.hypothesis(self.data,self.theta)
\n delta = prediction – self.labels
\n theta = self.theta
\n theta = theta – alpha*(1\/num_examples)*(np.dot(delta.T,self.data)).T
\n self.theta = theta<\/p>\n
def cost_function(self,data,labels):
\n """
\n \u635f\u5931\u8ba1\u7b97\u6a21\u5757
\n :param self:
\n :param data:
\n :param labels:
\n :return:
\n """
\n num_examples = data.shape[0]
\n delta = LinearRegression.hypothesis(self.data,self.theta) – labels
\n cost = (1\/2)*np.dot(delta.T,delta)\/num_examples
\n """print(cost.shape)"""
\n return cost[0][0]<\/p>\n
"""\u88c5\u9970\u5668"""
\n @staticmethod
\n def hypothesis(data,theta):
\n prediction = np.dot(data,theta)
\n return prediction<\/p>\n
def get_cost(self,data,labels):
\n data_processed = prepare_for_training(data,
\n self.polynomial_degree,
\n self.sinusoid_degree,
\n self.normalize_data
\n )[0]<\/p>\n
return self.cost_function(data_processed,labels)<\/p>\n
def predict(self,data):
\n data_processed = prepare_for_training(data,
\n self.polynomial_degree,
\n self.sinusoid_degree,
\n self.normalize_data
\n )[0]
\n predictions = LinearRegression.hypothesis(data_processed,self.theta)<\/p>\n
return predictions <\/p>\n
\u6587\u4ef62：Linear_regression.py\u00a0<\/p>\n
import numpy as np
\n"""\u7528\u4e8e\u79d1\u5b66\u8ba1\u7b97\u7684\u4e00\u4e2a\u5e93，\u63d0\u4f9b\u4e86\u591a\u7ef4\u6570\u7ec4\u5bf9\u8c61\u4ee5\u53ca\u64cd\u4f5c\u51fd\u6570"""
\nimport pandas as pd
\n"""\u4e00\u4e2a\u7528\u4e8e\u6570\u636e\u5bfc\u5165\u3001\u5bfc\u51fa\u3001\u6e05\u6d17\u548c\u5206\u6790\u7684\u5e93，\u672c\u6587\u4e2d\u5bfc\u5165csv\u683c\u5f0f\u6570\u636e\u7b49\u7b49"""
\nimport matplotlib.pyplot as plt
\n"""pyplot\u63d0\u4f9b\u4e86\u7ed8\u56fe\u63a5\u53e3"""
\nimport matplotlib
\n"""\u4e00\u4e2a\u5f3a\u5927\u7684\u7ed8\u56fe\u5e93"""<\/p>\n
# \u8bbe\u7f6ematplotlib\u6b63\u5e38\u663e\u793a\u4e2d\u6587\u548c\u8d1f\u53f7
\nmatplotlib.rcParams['font.family'] = 'SimHei' # \u6307\u5b9a\u9ed8\u8ba4\u5b57\u4f53\u4e3a\u9ed1\u4f53
\nmatplotlib.rcParams['axes.unicode_minus'] = False # \u6b63\u786e\u663e\u793a\u8d1f\u53f7<\/p>\n
from prepare_for_training import LinearRegression<\/p>\n
data = pd.read_csv("D:\/machine_learning\/archive\/2017.csv")
\ntrain_data = data.sample(frac = 0.8)
\ntest_data = data.drop(train_data.index)<\/p>\n
input_param_name = 'Economy..GDP.per.Capita.'
\noutput_param_name = 'Happiness.Score'<\/p>\n
x_train = train_data[[input_param_name]].values
\ny_train = train_data[[output_param_name]].values<\/p>\n
x_test = test_data[[input_param_name]].values
\ny_test = test_data[[output_param_name]].values<\/p>\n
plt.scatter(x_train,y_train,label ='Train data')
\nplt.scatter(x_test,y_test,label ='Test data')
\nplt.xlabel(input_param_name)
\nplt.ylabel(output_param_name)
\nplt.title('Happy')
\nplt.legend()
\nplt.show()<\/p>\n
"""\u8bad\u7ec3\u6b21\u6570，\u5b66\u4e60\u7387"""
\nnum_iterations = 500
\nlearning_rate = 0.01<\/p>\n
linear_regression = LinearRegression(x_train,y_train)
\n(theta,cost_history) = linear_regression.train(learning_rate,num_iterations)
\nprint('\u5f00\u59cb\u65f6\u7684\u635f\u5931',cost_history[0])
\nprint('\u8bad\u7ec3\u540e\u7684\u635f\u5931',cost_history[-1])<\/p>\n
plt.plot(range(num_iterations),cost_history)
\nplt.xlabel('Iter')
\nplt.ylabel('cost')
\nplt.title('\u635f\u5931\u503c')
\nplt.show()<\/p>\n
predictions_num = 100
\nx_predictions = np.linspace(x_train.min(),x_train.max(),predictions_num).reshape(predictions_num,1)
\ny_predictions = linear_regression.predict(x_predictions)<\/p>\n
plt.scatter(x_train,y_train,label ='Train data')
\nplt.scatter(x_test,y_test,label ='Test data')
\nplt.plot(x_predictions,y_predictions,'r',label = 'Prediction')
\nplt.xlabel(input_param_name)
\nplt.ylabel(output_param_name)
\nplt.title('Happy')
\nplt.legend()
\nplt.show()<\/p>\n
\u6548\u679c\u56fe： <\/p>\n
<\/p>\n
![\"\"](\"https:\/\/www.wsisp.com\/helps\/wp-content\/uploads\/2025\/05\/20250430220238-68129dfe811d1.png\")
<\/p>\n
![\"\"](\"https:\/\/www.wsisp.com\/helps\/wp-content\/uploads\/2025\/05\/20250430220238-68129dfe94533.png\")
<\/p>\n","protected":false},"excerpt":{"rendered":"
\u6587\u7ae0\u6d4f\u89c8\u9605\u8bfb1.5k\u6b21\uff0c\u70b9\u8d5e50\u6b21\uff0c\u6536\u85cf19\u6b21\u3002\u673a\u5668\u5b66\u4e60\uff0c\u7ebf\u6027\u56de\u5f52\u9884\u6d4b<\/p>\n","protected":false},"author":2,"featured_media":34804,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[1],"tags":[50,207,2949],"topic":[],"class_list":["post-34808","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-server","tag-50","tag-207","tag-2949"],"yoast_head":"\n