학술논문
Using 2-additive fuzzy measure in Multiple Classifier System
Document Type
Conference
Author
Source
2009 International Conference on Machine Learning and Cybernetics Machine Learning and Cybernetics, 2009 International Conference on. 2:877-880 Jul, 2009
Subject
Language
ISSN
2160-133X
2160-1348
2160-1348
Abstract
Fuzzy measure and integral are widely used in Multiple Classifier System (MCS). But the number of coefficients involved in the fuzzy integral model grows exponentially with the number of classifiers to be aggregated. The main difficulty is to identify all these coefficients. This paper does an attempt Using 2-additive fuzzy measure in Multiple Classifier System. Our conclusion is that when different interactions exist in different classifiers the complexity of the computation can be significantly reduced by 2-order additive measure.A simple example is included to illustrate the 2-order additive measure.