DocumentCode
2643044
Title
Towards online learning of a fuzzy classifier
Author
Visa, Sofia ; Ralescu, Anca
Author_Institution
Dept. of ECECS, Cincinnati Univ., OH, USA
fYear
2005
fDate
26-28 June 2005
Firstpage
557
Lastpage
561
Abstract
This study addresses issues related to the online applicability of a fuzzy classifier. In particular, it shows that a fuzzy classifier can be learned incrementally, and that in this process, imbalanced data sets, even when imbalance changes between classes can be used. Finally, it shows that for each class, examples and counter examples, can be effectively used. The most important aspect of the online fuzzy classifier is its perfect incremental aspect.
Keywords
fuzzy systems; learning (artificial intelligence); pattern recognition; adaptive learning system; fuzzy classifier; fuzzy modeling; imbalanced data set; incremental learning; online learning; pattern recognition; Adaptive systems; Computer applications; Counting circuits; Frequency; Fuzzy sets; Fuzzy systems; Learning systems; Pattern recognition; Testing; Training data; adaptive learning systems; fuzzy classifier; fuzzy modeling; online learning; pattern recognition;
fLanguage
English
Publisher
ieee
Conference_Titel
Fuzzy Information Processing Society, 2005. NAFIPS 2005. Annual Meeting of the North American
Print_ISBN
0-7803-9187-X
Type
conf
DOI
10.1109/NAFIPS.2005.1548596
Filename
1548596
Link To Document