import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
from sklearn import datasets
from sklearn import preprocessing
from sklearn.model_selection import train_test_split
from sklearn import linear_model
from sklearn.metrics import mean_squared_error,r2_score
import pickle
import warnings
warnings.filterwarnings('ignore')
data = datasets.load_boston()
x,y = data.data,data.target
standardized_x = preprocessing.StandardScaler().fit_transform(x)
x_train,x_test,y_train,y_test = train_test_split(standardized_x,y,test_size=0.2,random_state=1)
print('[Linear_Regression]')
clf = linear_model.LinearRegression(fit_intercept=True)
clf.fit(x_train,y_train)
print('模型(线性回归)参数')
print('intercept:\n',clf.intercept_)
print('coef:\n',clf.coef_)
#保存模型
with open('linear_regression.pkl','wb') as f:
pickle.dump(clf,f)
print('%%%model saving as linear_regression.pkl%%%')
#读取模型
with open('linear_regression.pkl','rb') as f:
model = pickle.load(f)
predict_train = clf.predict(x_train)
predict_test = clf.predict(x_test)
train_mse = round(np.sqrt(mean_squared_error(y_train,predict_train)),4)
train_r2 = round(r2_score(y_train,predict_train),2)
test_mse = round(np.sqrt(mean_squared_error(y_test,predict_test)),4)
test_r2 = round(r2_score(y_test,predict_test),2)
print('预测表现')
print('Train RMSE = ',train_mse,' Train R2 = ',train_r2)
print('Test RMSE = ',test_mse,' Test R2 = ',test_r2)
'''
岭回归(L2):引入正则化项来消除异常值
'''
print('\n[Ridge_Regression]')
clf = linear_model.Ridge(alpha=10,fit_intercept=True,max_iter=10000)
clf.fit(x_train,y_train)
print('模型(Ridge回归)参数')
print('intercept:\n',clf.intercept_)
print('coef:\n',clf.coef_)
#保存模型
with open('ridge_regression.pkl','wb') as f:
pickle.dump(clf,f)
print('%%%model saving as ridge_regression.pkl%%%')
#读取模型
with open('ridge_regression.pkl','rb') as f:
model = pickle.load(f)
predict_train = clf.predict(x_train)
predict_test = clf.predict(x_test)
train_mse = round(np.sqrt(mean_squared_error(y_train,predict_train)),4)
train_r2 = round(r2_score(y_train,predict_train),2)
test_mse = round(np.sqrt(mean_squared_error(y_test,predict_test)),4)
test_r2 = round(r2_score(y_test,predict_test),2)
print('预测表现')
print('Train RMSE = ',train_mse,' Train R2 = ',train_r2)
print('Test RMSE = ',test_mse,' Test R2 = ',test_r2)
'''
Lasso(L1):将某些系数压缩为0
'''
print('\n[Lasso_Regression]')
clf = linear_model.Lasso(alpha=0.01,fit_intercept=True,max_iter=10000)
clf.fit(x_train,y_train)
print('模型(Lasso回归)参数')
print('intercept:\n',clf.intercept_)
print('coef:\n',clf.coef_)
#保存模型
with open('lasso_regression.pkl','wb') as f:
pickle.dump(clf,f)
print('%%%model saving as lasso_regression.pkl%%%')
#读取模型
with open('lasso_regression.pkl','rb') as f:
model = pickle.load(f)
predict_train = clf.predict(x_train)
predict_test = clf.predict(x_test)
train_mse = round(np.sqrt(mean_squared_error(y_train,predict_train)),4)
train_r2 = round(r2_score(y_train,predict_train),2)
test_mse = round(np.sqrt(mean_squared_error(y_test,predict_test)),4)
test_r2 = round(r2_score(y_test,predict_test),2)
print('预测表现')
print('Train RMSE = ',train_mse,' Train R2 = ',train_r2)
print('Test RMSE = ',test_mse,' Test R2 = ',test_r2)