Alternating Direction Method of Multipliers for l_(1)-l_(2)-Regularized Logistic Regression Model  

作　　者：Yan-Qin Bai Kai-Ji Shen 

机构地区：[1]Department of Mathematics,Shanghai University,Shanghai 200444,China

出　　处：《Journal of the Operations Research Society of China》2016年第2期243-253,共11页中国运筹学会会刊（英文）

基　　金：the National Natural Science Foundation of China(No.11371242)。

摘　　要：Logistic regression has been proved as a promising method for machine learning,which focuses on the problem of classification.In this paper,we present anl_(1)-l_(2)-regularized logistic regression model,where thel1-norm is responsible for yielding a sparse logistic regression classifier and thel_(2)-norm for keeping betlter classification accuracy.To solve thel_(1)-l_(2)-regularized logistic regression model,we develop an alternating direction method of multipliers with embedding limitedlBroyden-Fletcher-Goldfarb-Shanno(L-BFGS)method.Furthermore,we implement our model for binary classification problems by using real data examples selected from the University of California,Irvine Machines Learning Repository(UCI Repository).We compare our numerical results with those obtained by the well-known LIBSVM and SVM-Light software.The numerical results show that ourl_(1)-l_(2)-regularized logisltic regression model achieves better classification and less CPU Time.

关 键 词：Classification problems Logistic regression model SPARSITY ALTERNATING direction method of multipliers 

分 类 号：O17[理学—数学]