Analysis of Rule Generation Techniques using Computational Method

Abstract

Problem statement: Enhanced decision making seeks a right of way mainly in business sectors for mining profitable patterns in databases. To address this problem, the authors have incorporated the temporal and utility concept of mining effective rules. Though the objective of generating profitable rule set is achieved. Still in effective decision making, Optimization of the rule set is required. Therefore the authors have used Particle Swarm Optimization technique for optimizing the generated rule set. To optimize the rule set, the authors have introduced weighted fitness function, which is based on the concept of weighted support and weighted confidence. Approach: In this article the authors have done a meticulous performance study of the proposed approach Optimized UTARM also called UTARM_PSO. Also, comparative analysis of the proposed approach with the apriori_PSO technique is done. Results and Conclusion: The experimental results show that the our approach has performed the optimization consistently and precisely on values of support ranging from 80-90%. The Comparative analysis pointed to a 15% in the overall performance of our approach against the existing approach.