岭回归（加入正则化的线性回归）的使用

coding=utf-8

“”"
author:lei
function: L2正则化 解决过拟合，使得w的每个元素都很小，都接近于0
越小的参数说明模型越简单，越简单的模型则越不容易产生过拟合现象
线性回归：容易出现过拟合，为了把训练集数据表现的更好，可使用L2正则化：Ridge：岭回归，带有正则化的线性回归，解决过拟合。
回归解决过拟合
“”"

岭回归

from sklearn.linear_model import Ridge
from sklearn.preprocessing import StandardScaler
from sklearn.metrics import mean_squared_error
from sklearn.model_selection import train_test_split
from sklearn.datasets import load_boston
from sklearn.model_selection import GridSearchCV

def ridge_regression():
rg = Ridge()

lb = load_boston()

x_train, x_test, y_train, y_test = train_test_split(lb.data, lb.target, test_size=0.25)

std_x = StandardScaler()
x_train = std_x.fit_transform(x_train)
x_test = std_x.transform(x_test)

std_y = StandardScaler()
y_train = std_y.fit_transform(y_train.reshape(-1, 1))
y_test = std_y.transform(y_test.reshape(-1, 1))

gc = GridSearchCV(rg, param_grid={"alpha": [0.5, 1, 0.8, 1.5]}, cv=5)

gc.fit(x_train, y_train)

y_predict = std_y.inverse_transform(gc.predict(x_test))

# print(y_predict)
print(gc.best_params_)
print(mean_squared_error(std_y.inverse_transform(y_test), y_predict))

if name == ‘main’:
ridge_regression()