An Improved Distributed Fuzzy Associative Classifier for Big Data Using Butterfly Optimization Algorithm based Artificial Bee Colony Algorithm

Abstract

Association rule mining is one of the primary tasks in data mining, which is helpful in knowing the intriguing associations among the elements in the item sets of a massive database. Association Rule Mining is a prominent data mining approach. Researchers have evaluated several factors of the approach; however, very less focus is paid to tackle with the reliability of the rules external to the dataset using which the generation of these rules is done. Aproiri constitutes the common algorithm of association rule mining that helps frequent item sets’ generation. Apriori utilizes minimal support threshold to get frequent items. In this work, an algorithm, formed by Butterfly Optimization Algorithm based Artificial Bee Colony Algorithm is presented which helps in choosing the association rules with Associative Classifiers (ACs). Rather than the ABC’s onlooker bee stage, the random walk process of BOA is utilized to improve the exploration. Butterfly Optimization Algorithm Based Artificial Bee Colony Algorithm (BOAABC) is used on the rules that the apriori algorithm generates, to choose the association rules. The validation process is carried out on datasets obtained from UCI database which reveal the performance of the proposed research work and the proposed technique is efficient in the choice of association rules rather than the available algorithms. In this work, it is confirmed that the rules created in the proposed research work are easy and understandable.