Adaptation in classification systems

This source preferred by Hamid Bouchachia

This data was imported from DBLP:

Authors: Bouchachia, A.

Editors: Hassanien, A.E., Abraham, A. and Herrera, F.

http://dx.doi.org/10.1007/978-3-642-01533-5

Volume: 202

Pages: 237-258

Publisher: Springer

ISBN: 978-3-642-01532-8

DOI: 10.1007/978-3-642-01533-5_9

This data was imported from Scopus:

Authors: Bouchachia, A.

Volume: 202

Pages: 237-258

ISBN: 9783642015328

DOI: 10.1007/978-3-642-01533-5_9

The persistence and evolution of systems essentially depend of their ability to self-adapt to new situations. As an expression of intelligence, adaptation is a distinguishing quality of any system that is able to learn and to adjust itself in a flexible manner to new environmental conditions. Such ability ensures selfcorrection over time as new events happen, new input becomes available, or new operational conditions occur. This requires self-monitoring of the performance in an ever changing environment. The relevance of adaptation is established in numerous domains and by versatile real world applications. The primary goal of this contribution is to investigate adaptation issues in learning classification systems formdifferent perspectives. Being a scheme of adaptation, life long incremental learning will be examined. However, special attention will be given to adaptive neural networks and the most visible incremental learning algorithms (fuzzy ARTMAP, nearest generalized exemplar, growing neural gas, generalized fuzzy minmax neural network, IL based on function decomposition) and their adaptation mechanisms will be discussed. Adaptation can also be incorporated in the combination of such incremental classifiers in different ways so that adaptive ensemble learners can be obtained too. These issues and other pertaining to drift will be investigated and illustrated by means of a numerical simulation. © 2009 Springer-Verlag Berlin Heidelberg.

The data on this page was last updated at 04:45 on September 21, 2017.