# -*- coding: utf-8 -*- ''' Created on 2018年1月17日 @author: Jason.F @summary: Scikit-Learn庫邏輯斯蒂L1正則化-特征選擇 ''' import pandas as pd import numpy as np from sklearn.cross_validation import train_test_split from sklearn.preprocessing import MinMaxScaler from sklearn.preprocessing import StandardScaler from sklearn.linear_model import LogisticRegression import matplotlib.pyplot as plt #導(dǎo)入數(shù)據(jù) df_wine = pd.read_csv('https://archive.ics.uci.edu/ml/machine-learning-databases/wine/wine.data',header=None) df_wine.columns=['Class label','Alcohol','Malic acid','Ash','Alcalinity of ash','Magnesium','Total phenols','Flavanoids','Nonflavanoid phenols','Proanthocyanins','Color intensity','Hue','OD280/OD315 of diluted wines','Proline'] print ('class labels:',np.unique(df_wine['Class label'])) #print (df_wine.head(5)) #分割訓(xùn)練集合測試集 X,y=df_wine.iloc[:,1:].values,df_wine.iloc[:,0].values X_train,X_test,y_train,y_test=train_test_split(X,y,test_size=0.3,random_state=0) #特征值縮放 #歸一化 mms=MinMaxScaler() X_train_norm=mms.fit_transform(X_train) X_test_norm=mms.fit_transform(X_test) #標(biāo)準(zhǔn)化 stdsc=StandardScaler() X_train_std=stdsc.fit_transform(X_train) X_test_std=stdsc.fit_transform(X_test)#L1正則化的邏輯斯蒂模型 lr=LogisticRegression(penalty='l1',C=0.1)#penalty='l2' lr.fit(X_train_std,y_train) print ('Training accuracy:',lr.score(X_train_std, y_train)) print ('Test accuracy:',lr.score(X_test_std, y_test))#比較訓(xùn)練集和測試集，觀察是否出現(xiàn)過擬合 print (lr.intercept_)#查看截距，三個(gè)類別 print (lr.coef_)#查看權(quán)重系數(shù)，L1有稀疏化效果做特征選擇#正則化效果，減少約束參數(shù)值C，增加懲罰力度，各特征權(quán)重系數(shù)趨近于0 fig=plt.figure() ax=plt.subplot(111) colors=['blue','green','red','cyan','magenta','yellow','black','pink','lightgreen','lightblue','gray','indigo','orange'] weights,params=[],[] for c in np.arange(-4,6,dtype=float):lr=LogisticRegression(penalty='l1',C=10**c,random_state=0)lr.fit(X_train_std,y_train)weights.append(lr.coef_[0])#三個(gè)類別，選擇第一個(gè)類別來觀察params.append(10**c) weights=np.array(weights) for column,color in zip(range(weights.shape[1]),colors):plt.plot(params,weights[:,column],label=df_wine.columns[column+1],color=color) plt.axhline(0,color='black',linestyle='--',linewidth=3) plt.xlim([10**(-5),10**5]) plt.ylabel('weight coefficient') plt.xlabel('C') plt.xscale('log') plt.legend(loc='upper left') ax.legend(loc='upper center',bbox_to_anchor=(1.38,1.03),ncol=1,fancybox=True) plt.show()

結(jié)果：

('class labels:', array([1, 2, 3], dtype=int64)) ('Training accuracy:', 0.9838709677419355) ('Test accuracy:', 0.98148148148148151) [-0.38378625 -0.15815556 -0.70033857] [[ 0.28028457 0. 0. -0.02806147 0. 0.0.71013567 0. 0. 0. 0. 0.1.23592372][-0.64368703 -0.06896342 -0.05715611 0. 0. 0. 0.0. 0. -0.92722893 0.05967934 0. -0.37098083][ 0. 0.06129709 0. 0. 0. 0.-0.63710764 0. 0. 0.49858959 -0.35822494 -0.570042510. ]]

《新程序員》：云原生和全面數(shù)字化實(shí)踐50位技術(shù)專家共同創(chuàng)作，文字、視頻、音頻交互閱讀

總結(jié)

以上是生活随笔為你收集整理的【Python-ML】SKlearn库L1正则化特征选择的全部內(nèi)容，希望文章能夠幫你解決所遇到的問題。

如果覺得生活随笔網(wǎng)站內(nèi)容還不錯(cuò)，歡迎將生活随笔推薦給好友。