Cooperative Coevolution for Learning Fuzzy Rule-Based Systems

Abstract

Abstract. In the last few years, the coevolutionary paradigm has shown an increasing interest thanks to its high ability to manage huge search spaces. Particularly, the cooperative interaction scheme is recommendable when the problem solution may be decomposable in subcomponents and there are strong interdependencies among them. The paper introduces a novel application of these algorithms to the learning of fuzzy rule-based systems for system modeling. Traditionally, this process is performed by sequentially designing their different components. However, we propose to accomplish a simultaneous learning process with cooperative coevolution to properly consider the tight relation among the components, thus obtaining more accurate models.

Assessing Hierarchical Cooperative CoEvolution

“…[1,8,14]). This is also the case with Pareto coevolutionary approaches [2,5,6,9] which mainly aim at balancing between different criteria in multi-objective optimization problems.…”

“…This is particularly the approach followed by coevolutionary algorithms which utilize separate populations to evolve partial entities of the problem [14,17]. In order to formulate a composite problem solution, individuals within different populations have to be selected, put together and operate in parallel [1,13]. Each population can use its own evolutionary parameters (e.g.…”

“…The majority of existing cooperative coevolutionary schemes have been applied in designing distributed structures consisting of smaller components [1,4,6,8,15]. However, these schemes are not able to sufficiently investigate hierarchically organized problems where simple components are parts of other more complex ones, formulating gradually more composite structures.…”

Coevolutionary Architectures With Straight Line Programs for Solving the Symbolic Regression Problem

“…Finally a solution of the problem is constructed by means of the combination of partial solutions obtained from the different populations. Some examples of application of cooperative coevolutionary strategies for solving problems can be found in (Wiegand et al, 2001) and (Casillas et al, 2006). This paper focuses on the design and the study of several coevolutionary strategies between Genetic Programming (GP) and Evolutionary Algorithms (EA).…”

Reference sharing: a new collaboration model for cooperative coevolution

“…CCEAs have been applied to a great number of optimization problems with varying success, including function optimization (Bucci and Pollack 2005;Potter and De Jong 1994), evolving artificial neural networks (García-Pedrajas et al 2003;Gomez 2003;Shi and Wu 2008), learning fuzzy systems (Casillas et al 2002;Pena-Reyes and Sipper 2000). A major characteristic of CCEAs is the ability to simplify the complexities of a problem through decomposition.…”