Implementation of bio-inspired algorithms in high utility itemset mining

Abstract

Utility based itemset mining hasevolved as an important research topic in data mining, having application in retail-market data analysis, stock market prediction, online advertising and so on. Bio-inspired computation attempts to replicate the way in which biological organisms and sub-organisms operate using abstract computing ideas from living phenomena or biological systems.This study focuses on the application of bio-inspired algorithms on high utility itemset mining. A detailed analysis on the performance of thesealgorithmswere conducted on various parameters such as execution time, memory usage and the number of high utility items identified. Experimental result suggest Particle Swarm Optimization excels in its efficiency in execution time and memory usage.When the number of high utility items identified are concerned, it is Genetic Algorithm which outperforms Particle Swarm optimization and Bats algorithm.