Full text
|
PDF
- Requires a PDF viewer, such as GSview, Xpdf or Adobe Acrobat Reader
Download (208kB) | Preview |
Flagged and Compact Fuzzy ART: Fuzzy ART in more efficient forms
Citation
Al Rawi, Kamal R. and Gonzalo Martín, Consuelo (2010). Flagged and Compact Fuzzy ART: Fuzzy ART in more efficient forms. "The International Journal of ACM Jordan", v. 1 (n. 3); pp. 98-117. ISSN 2078-7952.
Description
| Title: | Flagged and Compact Fuzzy ART: Fuzzy ART in more efficient forms |
|---|---|
| Author/s: |
|
| Item Type: | Article |
| Título de Revista/Publicación: | The International Journal of ACM Jordan |
| Date: | January 2010 |
| ISSN: | 2078-7952 |
| Volume: | 1 |
| Subjects: | |
| Freetext Keywords: | Compact fuzzy ART, Flagged fuzzy ART, ART ANN, Unsupervised ART, and Unsupervised learning |
| Faculty: | Facultad de Informática (UPM) |
| Department: | Arquitectura y Tecnología de Sistemas Informáticos |
| Creative Commons Licenses: | Recognition - No derivative works - Non commercial |
Abstract
Two new simplified algorithms for Fuzzy ART have been developed. Onlycommitted category nodes C rather than the full capacity of the category nodes N (N mayor que C) are involved in the determination of the winning categoy node. In addition tothat, the initialization for weights and choice values has been eliminated. This reduces a lot the training time without altering the categorazation accuracy.Although, the new architectures are presented toward the fuzzy ART ANN in this work. However, they can be applied to all module of ART.
More information
| Item ID: | 7260 |
|---|---|
| DC Identifier: | https://oa.upm.es/7260/ |
| OAI Identifier: | oai:oa.upm.es:7260 |
| Official URL: | http://iswsa2010.acm.org/ |
| Deposited by: | Memoria Investigacion |
| Deposited on: | 13 Jun 2011 09:26 |
| Last Modified: | 20 Apr 2016 16:26 |
