线性模型L1正则化——套索回归

Posted 2021-03-13 st-lovaer

 1 from sklearn.model_selection import train_test_split
 2 from sklearn.datasets import load_diabetes
 3 X,y=load_diabetes().data,load_diabetes().target
 4 X_train,X_test,y_train,y_test=train_test_split(X,y,random_state=8)
 5 
 6 from sklearn.linear_model import Lasso
 7 import numpy as np
 8 lasso=Lasso().fit(X_train,y_train)
 9 print("the coefficient:{}".format(lasso.coef_))
10 print(‘the intercept:{}‘.format(lasso.intercept_))
11 print("the score of this model:{:.3f}".format(lasso.score(X_test,y_test)))
12 print("the model uses {}".format(np.sum(lasso.coef_!=0))+" features
")

 1 lasso01=Lasso(alpha=0.1,max_iter=100000).fit(X_train,y_train)
 2 print("the coefficient:{}".format(lasso01.coef_))
 3 print(‘the intercept:{}‘.format(lasso01.intercept_))
 4 print("the score of this model:{:.3f}".format(lasso01.score(X_test,y_test)))
 5 print("the model uses {}".format(np.sum(lasso01.coef_!=0))+" features
")
 6 
 7 lasso001=Lasso(alpha=0.01,max_iter=100000).fit(X_train,y_train)
 8 print("the coefficient:{}".format(lasso001.coef_))
 9 print(‘the intercept:{}‘.format(lasso001.intercept_))
10 print("the score of this model:{:.3f}".format(lasso001.score(X_test,y_test)))
11 print("the model uses {}".format(np.sum(lasso001.coef_!=0))+" features
")
12 
13 lasso0001=Lasso(alpha=0.001,max_iter=100000).fit(X_train,y_train)
14 print("the coefficient:{}".format(lasso0001.coef_))
15 print(‘the intercept:{}‘.format(lasso0001.intercept_))
16 print("the score of this model:{:.3f}".format(lasso0001.score(X_test,y_test)))
17 print("the model uses {}".format(np.sum(lasso0001.coef_!=0))+" features
")
18 
19 lasso00001=Lasso(alpha=0.0001,max_iter=100000).fit(X_train,y_train)
20 print("the coefficient:{}".format(lasso00001.coef_))
21 print(‘the intercept:{}‘.format(lasso00001.intercept_))
22 print("the score of this model:{:.3f}".format(lasso00001.score(X_test,y_test)))
23 print("the model uses {}".format(np.sum(lasso00001.coef_!=0))+" features
")

 1 import matplotlib.pyplot as plt
 2 plt.plot(lasso.coef_,‘s‘,label=‘Lasso alpha=1‘)
 3 plt.plot(lasso01.coef_,‘^‘,label=‘Lasso alpha=0.1‘)
 4 plt.plot(lasso001.coef_,‘v‘,label=‘Lasso alpha=0.01‘)
 5 plt.plot(lasso0001.coef_,‘o‘,label=‘Lasso alpha=0.001‘)
 6 plt.plot(lasso00001.coef_,‘*‘,label=‘Lasso alpha=0.0001‘)
 7 plt.xlabel("coeffient index")
 8 plt.ylabel("coeffient magnitude")
 9 plt.legend(loc=(0,1.05))
10 plt.show()