Optimization of Association Rule Mining Using Hybridized Artificial Bee Colony (ABC) with BAT Algorithm

One of the major tasks of data mining is association rule mining, which is used for finding the interesting relationships among the items in itemsets of huge database. Aproiri is the familiar algorithm of association rule mining for generating frequent itemsets. Apriori uses minimum support threshold to find frequent items. In this paper, an algorithm called hybridization of ABC with BAT algorithm is proposed which is used for optimization of association rules. Instead of onlooker bee phase of ABC, random walk of BAT is used in order to increase the exploration. Hybridized ABC with BAT algorithm is applied on the rules generated from apriori algorithm, for optimizing association rules. The experiments are performed on datasets taken from UCI repository which show the proposed work performance and proposed methodology can effectively optimize association rules when compared to the existing algorithms. In the paper, we also proved that the rules generated using proposed work are simple and comprehensible.