Robust L1-norm multi-weight vector projection support vector machine with efficient algorithm

doi:10.1016/j.neucom.2018.04.083

Robust L1-norm multi-weight vector projection support vector machine with efficient algorithm

2018-11-01

15

PRE

AI

DOI：10.1016/j.neucom.2018.04.083

Original

Original request for help

Save

Abstract

En 中文

The recently proposed multi-weight vector projection support vector machine (EMVSVM) is an excellent multi-projections classifier. However, the formulation of MVSVM is based on the L-2-norm criterion, which makes it prone to be affected by outliers. To alleviate this issue, in this paper, we propose a robust L-1 norm MVSVM method, termed as MVSVM L-1. Specifically, our MVSVM L-1 aims to seek a pair of multiple projections such that, for each class, it maximizes the ratio of the L-1-norm between-class dispersion and the L-1-norm within-class dispersion. To optimize such L-1-norm ratio problem, a simple but efficient iterative algorithm is further presented. The convergence of the algorithm is also analyzed theoretically. Extensive experimental results on both synthetic and real-world datasets confirm the feasibility and effectiveness of the proposed MVSVM L-1. (c) 2018 Elsevier B.V. All rights reserved.

Keywords:

Support vector machine

Multi-weight vector projections

L-1-norm ratio optimization

Outliers

Multiple projections

Robust L1-norm multi-weight vector projection support vector machine with efficient algorithm

Abstract

AI Summary

Journal

Organization