Sparse Support Vector Machine with L _p Penalty for Feature Selection

详细信息    查看全文

• 作者：Lan Yao ; Feng Zeng ; Dong-Hui Li…
• 关键词：machine learning ; feature selection ; support vector machine ; Lp ; regularization
• 刊名：Journal of Computer Science and Technology
• 出版年：2017
• 出版时间：January 2017
• 年：2017
• 卷：32
• 期：1
• 页码：68-77
• 全文大小：
• 刊物类别：Computer Science
• 刊物主题：Computer Science, general; Software Engineering; Theory of Computation; Data Structures, Cryptology and Information Theory; Artificial Intelligence (incl. Robotics); Information Systems Applications (
• 出版者：Springer US
• ISSN：1860-4749
• 卷排序：32

文摘

We study the strategies in feature selection with sparse support vector machine (SVM). Recently, the socalled L_p-SVM (0 < p < 1) has attracted much attention because it can encourage better sparsity than the widely used L₁-SVM. However, L_p-SVM is a non-convex and non-Lipschitz optimization problem. Solving this problem numerically is challenging. In this paper, we reformulate the L_p-SVM into an optimization model with linear objective function and smooth constraints (LOSC-SVM) so that it can be solved by numerical methods for smooth constrained optimization. Our numerical experiments on artificial datasets show that LOSC-SVM (0 < p < 1) can improve the classification performance in both feature selection and classification by choosing a suitable parameter p. We also apply it to some real-life datasets and experimental results show that it is superior to L₁-SVM.