Publikacje

Abstract:

The paper presents an approach that integrates artificial neural networks, fuzzy logic, and - in supportive role - genetic algorithms, yielding a system that can be called a structured neuro-fuzzy (-genetic) classifier. In this approach, a fuzzy rule-based system is represented by a feed-forward network-like structure. It is able to learn and to generalize the gained knowledge as well as to explain (by providing a set of fuzzy rules describing the mechanisms of decision making) the decisions it makes. The proposed methodology has been successfully applied to the design of two decision support systems: for the diagnosing different types of the breast cancer and for the glass identification problem from the forensic science. A comparative analysis of the proposed methodology with several other approaches has also been performed.

POWRÓT

Strona główna > Publikacje

A neuro-fuzzy-genetic classifier for rule-based intelligent decision support