Mining Association Rules Using Adaptive Particle Swarm Optimization

Abstract

AbstractAssociation rule mining is a technique of data mining that is very widely used to deduce inferences from large databases. Particle swarm optimization is one of the several methods for mining association rules and has its own pitfalls. In this paper, an adaptive particle swarm optimization (APSO) that yields a finer solution by performing a diversified search over the entire search space is presented. The algorithmic parameters such as inertia weight and acceleration coefficients are adjusted dynamically to avoid possible local optima and to improve the convergence speed. The evolutionary state estimation (ESE) approach is adopted to identify the evolutionary states that the particle undergoes for each generation. The parameters are adjusted according to the estimated state in order to provide a better balance between global exploration and local exploitation. Additionally, an elitist learning strategy (ELS) is developed for the best particle to jump out of possible local optima. APSO provides a faster convergence mechanism and avoids premature convergence when compared to normal PSO.KeywordsParticle swarm optimizationAdaptive particle swarm optimizationEffectivenessAssociation rule mining