An Efficient Hybridization of Artificial Bee Colony with Levy distribution for Rare Itemsets Association Rule Mining
S. Selvarani, M. Jeyakarthic · 2020 Fourth International Conference on Inventive Systems and Control (ICISC) · 2020
Traditional association mining models depend on a support-confidence framework, allows the generation of frequent rules depending upon frequent itemsets identified. But, the infrequent itemsets called as rare itemsets are generally ignored by those techniques regularly hold meaningful information in certain reallife applications. In this paper, a new artificial bee colony (ABC) with Levy distribution called the ABC-L algorithm for the optimization of rare association rules. In the ABC algorithm, the exploration capability of the onlooker bee phase can be improved by the use of Levy distribution. The ABC-L algorithm is employed on the rules generated from the Apriori Rare (AR) algorithm to optimize the rare association results. The presented ARABC-L method is simulated using a set of three realtime synthetic datasets of T20D10000k, T100D10000k, T1000D10000k are used. The simulation results ensured that the rule generation process by the presented ABC-L is simple and comprehensible.