An experimental study of original and ordered fuzzy ARTMAP neural networks in pattern classification tasks

Kuan, Mei Ming, Lim, Chee Peng, Morad, Norhashimah and Harrison, Robert F. 2000, An experimental study of original and ordered fuzzy ARTMAP neural networks in pattern classification tasks, in TENCON 2000 : Proceedings : Intelligent systems and technologies for the new millennium, IEEE, Piscataway, N. J., pp. II-392-II-397.

Attached Files
Name Description MIMEType Size Downloads

Title An experimental study of original and ordered fuzzy ARTMAP neural networks in pattern classification tasks
Author(s) Kuan, Mei Ming
Lim, Chee Peng
Morad, Norhashimah
Harrison, Robert F.
Conference name Trends in Electronics Conference (2000 : Kuala Lumpur, Malaysia)
Conference location Kuala Lumpur, Malaysia
Conference dates 24-27 Sept. 2000
Title of proceedings TENCON 2000 : Proceedings : Intelligent systems and technologies for the new millennium
Editor(s) [Unknown]
Publication date 2000
Conference series Trends in Electronics Conference
Start page II-392
End page II-397
Total pages 6
Publisher IEEE
Place of publication Piscataway, N. J.
Keyword(s) adaptive resonance theory
fuzzy ARTMAP
max-min clustering
pattern classification
medical diagnosis
Summary This paper investigates the effectiveness of an ordering algorithm applied to the supervised Fuzzy ARTMAP (FAM) neural network in pattern classification tasks. Before presenting the input patterns to the FAM network (known as ordered FAM), a fixed order of input patterns is first identified using the ordering algorithm. An experimental study is conducted to compare the results from ordered FAM with the average and voting results from original FAM. In the study, a pool of the original FAM networks is trained using different sequences of input patterns, and the results are averaged. Outputs from various original FAM networks can also be combined using a majority voting strategy to reach a final result. A database comprising various symptoms and measurements of patients suffering from heart attack is used to evaluate the various schemes of the FAM network in medical pattern classification tasks. The results are compared, analyzed, and discussed.
ISBN 0780363558
Language eng
Field of Research 099999 Engineering not elsewhere classified
Socio Economic Objective 970109 Expanding Knowledge in Engineering
HERDC Research category E1.1 Full written paper - refereed
Persistent URL http://hdl.handle.net/10536/DRO/DU:30048768

Document type: Conference Paper
Collection: Institute for Frontier Materials
Connect to link resolver
 
Unless expressly stated otherwise, the copyright for items in DRO is owned by the author, with all rights reserved.

Versions
Version Filter Type
Access Statistics: 29 Abstract Views, 2 File Downloads  -  Detailed Statistics
Created: Wed, 26 Sep 2012, 09:25:34 EST

Every reasonable effort has been made to ensure that permission has been obtained for items included in DRO. If you believe that your rights have been infringed by this repository, please contact drosupport@deakin.edu.au.