A Genetic-Programming-Based Approach for the Learning of Compact Fuzzy Rule-Based Classification Systems

Abstract

Abstract. In the design of an interpretable fuzzy rule-based classification system (FRBCS) the precision as much as the simplicity of the extracted knowledge must be considered as objectives. In any inductive learning algorithm, when we deal with problems with a large number of features, the exponential growth of the fuzzy rule search space makes the learning process more difficult. Moreover it leads to an FRBCS with a rule base with a high cardinality. In this paper, we propose a genetic-programming-based method for the learning of an FRBCS, where disjunctive normal form (DNF) rules compete and cooperate among themselves in order to obtain an understandable and compact set of fuzzy rules, which presents a good classification performance with high dimensionality problems. This proposal uses a token competition mechanism to maintain the diversity of the population. The good results obtained with several classification problems support our proposal.

GP-COACH: Genetic Programming-based learning of COmpact and ACcurate fuzzy rule-based classification systems for High-dimensional problems

“…One example are those GFSs which use GP [53] for learning FRBSs. Most of these approaches follow the Pittsburgh codification scheme [1][2][3]35,43,55,60,62,68], although there are other proposals which use GCCL codification scheme [7,8,22,56].…”

A novel genetic cooperative-competitive fuzzy rule based learning method using genetic programming for high dimensional problems

“…In [21] Tsakonas investigates the effectiveness of GP-generated intelligent structures in classification tasks by means a context-free grammar that allows to encode several rules per individual in the population (Pittsburgh approach). Finally, the authors propose in [22] and [23] a GP-based algorithm for the learning of interpretable FRBCSs. The differences between this proposal and the one presented in this contribution will be later described in section IV.…”

GPFIS-CLASS: A Genetic Fuzzy System based on Genetic Programming for classification problems

“…This GFSs is guided to generate a compact fuzzy rule-based, codifying Disjunctive Normal Form (DNF) fuzzy rules in each Genetic Programming individual in a Genetic CooperativeCompetitive Learning style. • FRBCS-GP: Berlanga et al[12] presented a GFS very similar to GP-COACH, but not considering some genetic operators and population dynamics.…”