arrow
Return

A projection multi-objective SVM method for multi-class classification

delete2021-08-01
delete10
delete
OA
AI
Ling Liu cover
Ling Liu (Ling Liu) *
B
Belén Martín-Barragán
F
Francisco J. Prieto
DOI:10.1016/j.cie.2021.107425delete
deleteOriginal
deleteShare
deleteSave
View PDF
Abstract

Abstract

En 中文
Support Vector Machines (SVMs), originally proposed for classifications of two classes, have become a very popular technique in the machine learning field. For multi-class classifications, various single-objective models and multi-objective ones have been proposed. However,in most single-objective models, neither the different costs of different misclassifications nor the users' preferences were considered. This drawback has been taken into account in multi-objective models. In these models, large and hard second-order cone programs(SOCPs) were constructed ane weakly Pareto-optimal solutions were offered. In this paper, we propose a Projected Multi-objective SVM (PM), which is a multi-objective technique that works in a higher dimensional space than the object space. For PM, we can characterize the associated Pareto-optimal solutions. Additionally, it significantly alleviates the computational bottlenecks for classifications with large numbers of classes. From our experimental results, we can see PM outperforms the single-objective multi-class SVMs (based on an all-together method, one-against-all method and one-against-one method) and other multi-objective SVMs. Compared to the single-objective multi-class SVMs, PM provides a wider set of options designed for different misclassifications, without sacrificing training time. Compared to other multi-objective methods, PM promises the out-of-sample quality of the approximation of the Pareto frontier, with a considerable reduction of the computational burden.
Keywords:
Multiple objective programming
Support vector machine
Multi-class multi-objective SVM
Pareto-optimal solution

Journal

Computers and Industrial Engineering cover
Computers and Industrial Engineering
IF:
6.5
Papers:
1.0W
Citations:
4.8W

Organization

U
Universidad Carlos III de Madrid
Scholars:
207
Papers: 103
Citations: 2.5K
Beijing University of Technology cover
Beijing University of Technology
Scholars:
4.0K
Papers: 1.2K
Citations: 1.8W
U
University of Edinburgh
Scholars:
2.9K
Papers: 1.1K
Citations: 3.7W
researcher View more organizations
Cited Papers

Cited Papers

err
IF0
err
err0
PREAI
err
errShare
errSave
An energy-aware routing protocol for wireless sensor network based on genetic algorithm
err2017-06-22
err0
PREAI
errLingping Kong; Jeng-Shyang Pan; Václav Snášel; Pei-Wei Tsai; Tien-Wen Sung
errShare
errSave
Gene selection for cancer classification using support vector machines
err2002-01-01
err7.5K
errOAAI
errGuyon, I; Weston, J; Barnhill, S; Vapnik, V
errShare
errSave
Surface potential and magnetic properties of La0.7Sr0.3MnO3 periodic arrays fabricated by direct electron beam writing
err2008-07-29
err0
PREAI
errMing-Chung Wu; Yi-Jen Wu; Yu-Ching Huang; Chih-Min Chuang; Kuo-Chung Cheng; Ching-Fuh Lin; Yang-Fang Chen; Wei-Fang Su
errShare
errSave
A survey of recent developments in multiobjective optimization
err2007-06-01
err270
PREAI
errChinchuluun, Altannar; Pardalos, Panos M.
errShare
errSave
Saving phase: Injectivity and stability for phase retrieval
err2014-07-01
err0
errOAAI
errAfonso S. Bandeira; Jameson Cahill; Dustin G. Mixon; Aaron A. Nelson
errShare
errSave
researcher View more