Data dimensionality reduction with application to simplifying RBF network structure and improving classification performance

For high dimensional data, if no preprocessing is carried out before inputting patterns to classifiers, the computation required may be too heavy. For example, the number of hidden units of a radial basis function (RBF) neural network can be too large. This is not suitable for some practical applica...

Full description

Saved in:
Bibliographic Details
Main Authors: Wang, Lipo., Fu, Xiuju
Other Authors: School of Electrical and Electronic Engineering
Format: Article
Language:English
Published: 2012
Subjects:
Online Access:https://hdl.handle.net/10356/93971
http://hdl.handle.net/10220/8196
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Nanyang Technological University
Language: English
id sg-ntu-dr.10356-93971
record_format dspace
spelling sg-ntu-dr.10356-939712020-03-07T14:02:41Z Data dimensionality reduction with application to simplifying RBF network structure and improving classification performance Wang, Lipo. Fu, Xiuju School of Electrical and Electronic Engineering DRNTU::Engineering::Electrical and electronic engineering For high dimensional data, if no preprocessing is carried out before inputting patterns to classifiers, the computation required may be too heavy. For example, the number of hidden units of a radial basis function (RBF) neural network can be too large. This is not suitable for some practical applications due to speed and memory constraints. In many cases, some attributes are not relevant to concepts in the data at all. In this paper, we propose a novel separability-correlation measure (SCM) to rank the importance of attributes. According to the attribute ranking results, different attribute subsets are used as inputs to a classifier, such as an RBF neural network. Those attributes that increase the validation error are deemed irrelevant and are deleted. The complexity of the classifier can thus be reduced and its classification performance improved. Computer simulations show that our method for attribute importance ranking leads to smaller attribute subsets with higher accuracies compared with the existing SUD and Relief-F methods. We also propose a modified method for efficient construction of an RBF classifier. In this method we allow for large overlaps between clusters corresponding to the same class label. Our approach significantly reduces the structural complexity of the RBF network and improves the classification performance. Accepted version 2012-06-12T06:41:58Z 2019-12-06T18:48:38Z 2012-06-12T06:41:58Z 2019-12-06T18:48:38Z 2003 2003 Journal Article Fu, X., & Wang, L. (2003). Data dimensionality reduction with application to simplifying RBF network structure and improving classification performance. IEEE Transactions on Systems, Man, and Cybernetics – Part B: Cybernetics, 33(3), 399-409. https://hdl.handle.net/10356/93971 http://hdl.handle.net/10220/8196 10.1109/TSMCB.2003.810911 en IEEE transactions on systems, man, and cybernetics – Part B: cybernetics © 2003 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works. The published version is available at: [http://dx.doi.org/10.1109/TSMCB.2003.810911]. 11 p. application/pdf
institution Nanyang Technological University
building NTU Library
country Singapore
collection DR-NTU
language English
topic DRNTU::Engineering::Electrical and electronic engineering
spellingShingle DRNTU::Engineering::Electrical and electronic engineering
Wang, Lipo.
Fu, Xiuju
Data dimensionality reduction with application to simplifying RBF network structure and improving classification performance
description For high dimensional data, if no preprocessing is carried out before inputting patterns to classifiers, the computation required may be too heavy. For example, the number of hidden units of a radial basis function (RBF) neural network can be too large. This is not suitable for some practical applications due to speed and memory constraints. In many cases, some attributes are not relevant to concepts in the data at all. In this paper, we propose a novel separability-correlation measure (SCM) to rank the importance of attributes. According to the attribute ranking results, different attribute subsets are used as inputs to a classifier, such as an RBF neural network. Those attributes that increase the validation error are deemed irrelevant and are deleted. The complexity of the classifier can thus be reduced and its classification performance improved. Computer simulations show that our method for attribute importance ranking leads to smaller attribute subsets with higher accuracies compared with the existing SUD and Relief-F methods. We also propose a modified method for efficient construction of an RBF classifier. In this method we allow for large overlaps between clusters corresponding to the same class label. Our approach significantly reduces the structural complexity of the RBF network and improves the classification performance.
author2 School of Electrical and Electronic Engineering
author_facet School of Electrical and Electronic Engineering
Wang, Lipo.
Fu, Xiuju
format Article
author Wang, Lipo.
Fu, Xiuju
author_sort Wang, Lipo.
title Data dimensionality reduction with application to simplifying RBF network structure and improving classification performance
title_short Data dimensionality reduction with application to simplifying RBF network structure and improving classification performance
title_full Data dimensionality reduction with application to simplifying RBF network structure and improving classification performance
title_fullStr Data dimensionality reduction with application to simplifying RBF network structure and improving classification performance
title_full_unstemmed Data dimensionality reduction with application to simplifying RBF network structure and improving classification performance
title_sort data dimensionality reduction with application to simplifying rbf network structure and improving classification performance
publishDate 2012
url https://hdl.handle.net/10356/93971
http://hdl.handle.net/10220/8196
_version_ 1681048511924142080