线性回归，岭回归，lasso回归模型代码

import sklearn
import numpy as np
from sklearn import datasets, linear_model
import matplotlib.pyplot as plt
from sklearn.model_selection import train_test_split
print(sklearn.__version__)

0.20.3

定义训练和测试函数，训练函数输出拟合误差，测试函数返回预测结果与预测得分
# 模型回归结果
def trainLinearRegression(X, y):
    model = linear_model.LinearRegression()
    model.fit(X,y)
    print('E(|y-wTx)|^2):' + str(np.mean((model.predict(X)-y)**2)))
    return model 
# 岭回归
def trainRidgeRegression(X,y,coeff):
    model = linear_model.Ridge(alpha = coeff)
    model.fit(X,y)
    print('E(|y-wTx|^2):' + str(np.mean((model.predict(X) - y)**2)))
    return model 
# lasso回归
def trainLassoRegression(X,y,coeff):
    model = linear_model.Lasso(alpha = coeff)
    model.fit(X,y)
    print('E(|y-wTx|^2):' + str(np.mean((model.predict(X) - y)**2)))
    return model
# 模型测试得分
def testModel(X_test, y_test, model):
    score = model.score(X_test, y_test)
    y_pred = model.predict(X_test)
    print('model test score is '+str(score))
    return y_pred, score

利用数据波士顿房价预测
data, label = datasets.load_boston(return_X_y = True)
print(data.shape, label.shape)

(506, 13) (506,)

#数据切分
X_train, X_test, y_train, y_test = train_test_split(data, label, test_size = 0.2)
print(X_train.shape, y_train.shape, X_test.shape, y_test.shape)

(404, 13) (404,) (102, 13) (102,)

linearModel = trainLinearRegression(X_train, y_train)
ridgeModel = trainRidgeRegression(X_train, y_train, 0.5)
lassoModel = trainLassoRegression(X_train, y_train, 0.01)

E(|y-wTx)|^2):23.035621901695357
E(|y-wTx|^2):23.106381821638468
E(|y-wTx|^2):23.072139176720263

结果可知：线性拟合较好，lasso次之，岭回归拟合相对最差，可能是有了正则化，而牺牲了精度

y_pred_linear, score_linear = testModel(X_test, y_test, linearModel)
y_pred_ridge, score_ridge = testModel(X_test, y_test, ridgeModel)
y_pred_lasso, score_lasso = testModel(X_test, y_test, lassoModel)

model test score is 0.711048675536282
model test score is 0.7073389448825249
model test score is 0.7092400405765255

结果可知：从测试集上看，线性预测最好，岭回归次之，而lasso变成最差

plt.figure()
plt.plot(range(30), y_pred_linear[:30], 'bo-', y_test[:30], 'y.--')
plt.title('linear model prediction', fontsize = 'large')
plt.figure()
plt.plot(range(30), y_pred_ridge[:30], 'r>-', y_test[:30], 'y.--')
plt.title('linear model prediction', fontsize = 'large')
plt.figure()
plt.plot(range(30), y_pred_lasso[:30], 'ks-', y_test[:30], 'y.--')
plt.title('linear model prediction', fontsize = 'large')
plt.show()