Enhancing the fuzzy modeling tool GUAJE with a new module for fingrams-based analysis of fuzzy rule bases

Alonso, José María; Pancho, David P.; Magdalena, L.

Year: 2012

Type of Publication: In Proceedings

Keywords: fuzzy modeling; free software; social network analysis

Book title: IEEE International Conference on Fuzzy Systems, World Congress on Computational Intelligence (WCCI)

Pages: 1082-1089

Address: Brisbane (Australia)

Month: June

ISBN: 978-1-4673-1505-0

Abstract:

The so-called fuzzy inference-grams (fingrams) constitute a novel and powerful tool for expert analysis of fuzzy rule bases at inference level in terms of co-fired rules, i.e., rules fired at the same time by a given input vector. A fuzzy rule base can be seen as a population made up of a set of individual rules which are competing and collaborating among them with the aim of yielding both good generality-specificity and interpretabilityaccuracy trade-offs. Thus, system behavior turns up from the existing relations among rules which can be easily analyzed by looking at the rule base as a fingram, i.e., as a social network made of nodes representing fuzzy rules. This paper introduces a new software module for fingram generation and analysis which is provided with the last version of GUAJE, a free software tool devoted to the generation of understandable and accurate fuzzy systems. Notice that, fingram analysis consists of studying the interaction among nodes in the network for the purpose of understanding the structure and behavior of the fuzzy rule base under consideration. It is based on the basic principles of social network analysis which have been properly incorporated into GUAJE and adapted to the design of fuzzy systems.