A Maximum Class Distance Support Vector Machine-Based Algorithm for Recursive Dimension Reduction

Authors:
Zheng Sun;Xiaoguang Zhang;Dianxu Ruan;Guiyun Xu
Affiliations:
College of Mechanical and Electrical Engineering, China University of Mining and Technology, Xuzhou, China 221116;College of Mechanical and Electrical Engineering, China University of Mining and Technology, Xuzhou, China 221116 and Department of Electronic Science and Engineering, Nanjing University, Nanjing, ...;College of Mechanical and Electrical Engineering, China University of Mining and Technology, Xuzhou, China 221116;College of Mechanical and Electrical Engineering, China University of Mining and Technology, Xuzhou, China 221116
Venue:
ISNN 2009 Proceedings of the 6th International Symposium on Neural Networks: Advances in Neural Networks - Part II
Year:
2009

Citing 7
Cited 0

The nature of statistical learning theory

The nature of statistical learning theory
Variable selection in kernel Fisher discriminant analysis by means of recursive feature elimination

Computational Statistics & Data Analysis
Identification of contributing variables using kernel-based discriminant modeling and reconstruction

Expert Systems with Applications: An International Journal
A rough margin based support vector machine

Information Sciences: an International Journal
A neuro-fuzzy scheme for simultaneous feature selection and fuzzy rule-based classification

IEEE Transactions on Neural Networks
Recursive Support Vector Machines for Dimensionality Reduction

IEEE Transactions on Neural Networks
Performing Feature Selection With Multilayer Perceptrons

IEEE Transactions on Neural Networks

Quantified Score

Hi-index	0.00

Visualization

Abstract

A maximum class distance support vector machine based on the recursive dimension reduction is proposed. This algorithm referring to the concept of fisher linear discriminate analysis is introduced to make the distance between the classes as long as possible along the direction of the discriminate vector, and at the same time a classification hyper-plane with the largest distance between the two classes is achieved. Thus the classification hyper-plane can effectively consist with the distribution of samples, resulting to higher classification accuracy. This paper presents the recursive dimension reduction algorithm and its details. Finally, a simulation illustrates the effectiveness of the presented algorithm.