Francisco Herrera

Analysis and guidelines to obtain a good uniform fuzzy partition granularity for fuzzy rule-based systems using simulated annealing☆

Abstract In this contribution, we will analyse the importance of the fuzzy partition granularity for the linguistic variables in the design of fuzzy rule-based systems (FRBSs). In order to put this into effect, we will study the FRBS behaviour considering uniform fuzzy partitions with the same number of labels for all the linguistic variables, and considering uniform fuzzy partitions with any number of labels for each linguistic variable. We will present a method based on Simulated Annealing (SA) in order to obtain a good uniform fuzzy partition granularity that improves the FRBS behaviour. It is an efficient granularity search method for finding a good number of labels per variable.

A Short Study on the Use of Genetic 2-Tuples Tuning for Fuzzy Rule Based Classification Systems in Imbalanced Data-Sets

In this work our aim is to increase the performance of fuzzy rule based classifications systems in the framework of imbalanced data-sets by means of the application of a genetic tuning step. We focus on the imbalanced data-set problem since it appears in many real application areas and, for this reason, it has become a relevant topic in the area of machine learning. This problem occurs when the number of examples that represents one of the concepts of interest (usually the most important) is much lower than that of the remaining ones. We want to adapt the 2-tuples based genetic tuning approach to classification problems and to study the positive synergy between this method and the Chi et al.s fuzzy learning method, which is a basic approach in order to build the initial knowledge base. The experimental results show the improvement achieved by the 2-tuples based genetic tuning over the fuzzy rule based classification system in all types of imbalanced data, obtaining a better behaviour than the basic approach.