An Analysis Of Two Dimensionality Reduction Techniques On The Performance Of Neural Network Classifiers

This project involves an analysis of the effectiveness of two dimensionality reduction techniques, i.e., Principal Component Analysis as the standard approach and Random Projection as a recent technique. The study is based on the performance of two supervised neural network classifiers i.e., S...

Full description

Bibliographic Details
Main Author: Ong, Siok Lan
Format: Monograph
Language:English
Published: Universiti Sains Malaysia 2005
Subjects:
Online Access:http://eprints.usm.my/57588/
http://eprints.usm.my/57588/1/An%20Analysis%20Of%20Two%20Dimensionality%20Reduction%20Techniques%20On%20The%20Performance%20Of%20Neural%20Network%20Classifiers_Ong%20Siok%20Lan.pdf
_version_ 1848883664051503104
author Ong, Siok Lan
author_facet Ong, Siok Lan
author_sort Ong, Siok Lan
building USM Institutional Repository
collection Online Access
description This project involves an analysis of the effectiveness of two dimensionality reduction techniques, i.e., Principal Component Analysis as the standard approach and Random Projection as a recent technique. The study is based on the performance of two supervised neural network classifiers i.e., Standard Backpropagation and Fuzzy ARTMAP. A set of benchmark and real medical databases are used to evaluate the performance of the neural network models. The performance indicators used are percentage of correct classification, purity, and collective entropy. The Student’s two-tailed paired t-test is used to compare the significance of differences of the results. Based on the estimated 95% confidence intervals, a strong decision which eventually leads to a convincing conclusion on the performance of the dimensionality reduction techniques can be obtained. The perceived experimental results especially from the real medical data sets are encouraging enough to prove that Random Projection exhibits good performance as a dimensionality reduction technique. Surprisingly, Random Projection is effective on low dimensional data, and the outcomes are as good as Principal Component Analysis. A discussion on generalization of the results obtained is included, and a conclusion ensues. Recommendations are also included for further improvements and enhancements in the analysis of dimensionality reduction techniques.
first_indexed 2025-11-15T18:54:24Z
format Monograph
id usm-57588
institution Universiti Sains Malaysia
institution_category Local University
language English
last_indexed 2025-11-15T18:54:24Z
publishDate 2005
publisher Universiti Sains Malaysia
recordtype eprints
repository_type Digital Repository
spelling usm-575882023-03-28T06:10:25Z http://eprints.usm.my/57588/ An Analysis Of Two Dimensionality Reduction Techniques On The Performance Of Neural Network Classifiers Ong, Siok Lan T Technology TK Electrical Engineering. Electronics. Nuclear Engineering This project involves an analysis of the effectiveness of two dimensionality reduction techniques, i.e., Principal Component Analysis as the standard approach and Random Projection as a recent technique. The study is based on the performance of two supervised neural network classifiers i.e., Standard Backpropagation and Fuzzy ARTMAP. A set of benchmark and real medical databases are used to evaluate the performance of the neural network models. The performance indicators used are percentage of correct classification, purity, and collective entropy. The Student’s two-tailed paired t-test is used to compare the significance of differences of the results. Based on the estimated 95% confidence intervals, a strong decision which eventually leads to a convincing conclusion on the performance of the dimensionality reduction techniques can be obtained. The perceived experimental results especially from the real medical data sets are encouraging enough to prove that Random Projection exhibits good performance as a dimensionality reduction technique. Surprisingly, Random Projection is effective on low dimensional data, and the outcomes are as good as Principal Component Analysis. A discussion on generalization of the results obtained is included, and a conclusion ensues. Recommendations are also included for further improvements and enhancements in the analysis of dimensionality reduction techniques. Universiti Sains Malaysia 2005-03-01 Monograph NonPeerReviewed application/pdf en http://eprints.usm.my/57588/1/An%20Analysis%20Of%20Two%20Dimensionality%20Reduction%20Techniques%20On%20The%20Performance%20Of%20Neural%20Network%20Classifiers_Ong%20Siok%20Lan.pdf Ong, Siok Lan (2005) An Analysis Of Two Dimensionality Reduction Techniques On The Performance Of Neural Network Classifiers. Project Report. Universiti Sains Malaysia, Pusat Pengajian Kejuruteraan Elektrik dan Elektronik. (Submitted)
spellingShingle T Technology
TK Electrical Engineering. Electronics. Nuclear Engineering
Ong, Siok Lan
An Analysis Of Two Dimensionality Reduction Techniques On The Performance Of Neural Network Classifiers
title An Analysis Of Two Dimensionality Reduction Techniques On The Performance Of Neural Network Classifiers
title_full An Analysis Of Two Dimensionality Reduction Techniques On The Performance Of Neural Network Classifiers
title_fullStr An Analysis Of Two Dimensionality Reduction Techniques On The Performance Of Neural Network Classifiers
title_full_unstemmed An Analysis Of Two Dimensionality Reduction Techniques On The Performance Of Neural Network Classifiers
title_short An Analysis Of Two Dimensionality Reduction Techniques On The Performance Of Neural Network Classifiers
title_sort analysis of two dimensionality reduction techniques on the performance of neural network classifiers
topic T Technology
TK Electrical Engineering. Electronics. Nuclear Engineering
url http://eprints.usm.my/57588/
http://eprints.usm.my/57588/1/An%20Analysis%20Of%20Two%20Dimensionality%20Reduction%20Techniques%20On%20The%20Performance%20Of%20Neural%20Network%20Classifiers_Ong%20Siok%20Lan.pdf