The Rule Extraction of Numerical Association Rule Mining Using Hybrid Evolutionary Algorithm

Abstract

The topic of Particle Swarm Optimization (PSO) has recently gained popularity. Researchers has used it to solve difficulties related to job scheduling, evaluation of stock markets and association rule mining optimization. However, the PSO method often encounters the problem of getting trapped in the local optimum. Some researchers proposed a solution to over come that problem using combination of PSO and Cauchy distribution because this performance proved to reach the optimal rules. In this paper, we focus to adopt the combination for solving association rule mining (ARM) optimization problem in numerical dataset. Therefore, the aim of this research is to extract the rule of numerical ARM optimization problem for certain multi-objective functions such as support, confidence, and amplitude. This method is called PARCD. It means that PSO for numerical association rule mining problem with Cauchy Distribution. PARCD performed better results than other methods such as MOPAR, MODENAR, GAR, MOGAR and RPSOA.