Sabokrou M.,University of valie asr |
Fallahi N.,Shiraz University |
Hoseini M.,Amirkabir University of Technology
International Review on Computers and Software | Year: 2012
In fuzzy rule-based classification systems (FRBCS), a set of rules for classification is used and to achieve more accuracy, weight is used for these rules. In the past researchers proposed some heuristic methods to produce and to weight fuzzy rules. This paper uses Evolutionary Strategy as a method based on population gradual adjustment by environmental conditions that can select a few simple fuzzy if-then rules for pattern classification problems with many continuous features and calculate efficient weighs for these rules. The proposed method has been compared with existing fuzzy classification rule weighing methods. this is done through computer simulations on some well-known classification problems with continuous features. This comparison shows that our solution is better than the other methods. © 2012 Praise Worthy Prize S.r.l. - All rights reserved.