A Weighted Particle Swarm Optimization Technique for Optimizing Association Rules

Abstract

The Process of finding out correlations among the data items in the databases forms the core concept in Association Rule Mining. The Association rule Mining algorithms helps in decision making process. Since it plays a vital role in this area, the rules generated by the algorithms should be of less in number and precise. Association rule algorithms, such as Apriori, scrutinize a long list of transactions in order to decide which items are most commonly purchased together. The challenge of digging out association patterns from data draws upon research in databases, machine learning and optimization to bring advanced intelligent solutions. But even though it provides some robustness, the rules generated from the algorithm may be redundant in some cases. So in order to overcome the problems we need to optimize the rules generated from these algorithms. Here we consider the Utility based Temporal Association Rule Mining method for generating the association rules and the Particle Swam Optimization algorithm is used to optimize the generated rules. The main processes in this proposed approach are calculation of the support and confidence from the input data, the Rule generation, Initialization , updation of the velocity , position of the rules and evaluation of fitness function. This paper attempts to use the PSO technique to optimize the utility based temporal rules by filtering out the redundant rules and thereby reducing the problem space.