机器学习之路：python k近邻回归 预测波士顿房价

             1
            from sklearn.datasets import load_boston
 2from sklearn.cross_validation import train_test_split
 3from sklearn.preprocessing import StandardScaler
 4from sklearn.neighbors import KNeighborsRegressor
 5from sklearn.metrics import r2_score, mean_squared_error, mean_absolute_error
 6import numpy as np
 7 8# 1 准备数据 9# 读取波士顿地区房价信息10 boston = load_boston()
11# 查看数据描述12# print(boston.DESCR)   # 共506条波士顿地区房价信息，每条13项数值特征描述和目标房价13# 查看数据的差异情况14# print("最大房价：", np.max(boston.target))   # 5015# print("最小房价：",np.min(boston.target))    # 516# print("平均房价：", np.mean(boston.target))   # 22.5328063241106771718 x = boston.data
19 y = boston.target
2021# 2 分割训练数据和测试数据22# 随机采样25%作为测试 75%作为训练23 x_train, x_test, y_train, y_test = train_test_split(x, y, test_size=0.25, random_state=33)
242526# 3 训练数据和测试数据进行标准化处理27 ss_x = StandardScaler()
28 x_train = ss_x.fit_transform(x_train)
29 x_test = ss_x.transform(x_test)
3031 ss_y = StandardScaler()
32 y_train = ss_y.fit_transform(y_train.reshape(-1, 1))
33 y_test = ss_y.transform(y_test.reshape(-1, 1))
3435# 4 两种k近邻回归行学习和预测36# 初始化k近邻回归模型 使用平均回归进行预测37 uni_knr = KNeighborsRegressor(weights="uniform")
38# 训练39uni_knr.fit(x_train, y_train)
40# 预测 保存预测结果41 uni_knr_y_predict = uni_knr.predict(x_test)
4243# 多初始化k近邻回归模型 使用距离加权回归44 dis_knr = KNeighborsRegressor(weights="distance")
45# 训练46dis_knr.fit(x_train, y_train)
47# 预测 保存预测结果48 dis_knr_y_predict = dis_knr.predict(x_test)
4950# 5 模型评估51# 平均k近邻回归 模型评估52print("平均k近邻回归的默认评估值为：", uni_knr.score(x_test, y_test))
53print("平均k近邻回归的R_squared值为：", r2_score(y_test, uni_knr_y_predict))
54print("平均k近邻回归的均方误差为:", mean_squared_error(ss_y.inverse_transform(y_test),
55                                              ss_y.inverse_transform(uni_knr_y_predict)))
56print("平均k近邻回归 的平均绝对误差为:", mean_absolute_error(ss_y.inverse_transform(y_test),
57                                                 ss_y.inverse_transform(uni_knr_y_predict)))
58# 距离加权k近邻回归 模型评估59print("距离加权k近邻回归的默认评估值为：", dis_knr.score(x_test, y_test))
60print("距离加权k近邻回归的R_squared值为：", r2_score(y_test, dis_knr_y_predict))
61print("距离加权k近邻回归的均方误差为:", mean_squared_error(ss_y.inverse_transform(y_test),
62                                           ss_y.inverse_transform(dis_knr_y_predict)))
63print("距离加权k近邻回归的平均绝对误差为:", mean_absolute_error(ss_y.inverse_transform(y_test),
64                                              ss_y.inverse_transform(dis_knr_y_predict)))
6566‘‘‘67平均k近邻回归的默认评估值为： 0.6903454564606561
68平均k近邻回归的R_squared值为： 0.6903454564606561
69平均k近邻回归的均方误差为: 24.01101417322835
70平均k近邻回归 的平均绝对误差为: 2.9680314960629928
71距离加权k近邻回归的默认评估值为： 0.7197589970156353
72距离加权k近邻回归的R_squared值为： 0.7197589970156353
73距离加权k近邻回归的均方误差为: 21.730250160926044
74距离加权k近邻回归的平均绝对误差为: 2.8050568785108005
75‘‘‘