¬High Utility Itemsets Mining for Improved Particle Swarm Optimization (IPSO)

Abstract

 Nowadays, data mining and knowledge discovery applications extend its focus on Mining High Utility Itemsets (HUI). Discovery of relevant optimum values in every itemsets slong utilities no lesser than low threshold value are done by ­High Utility Itemsets Mining(HUIM). Previously, Bio-HUI framework based algorithms be designed on supporting genetic algorithm, particle swarm optimization, and the bat algorithm for discovery of HUIs. roulette wheel selection process accordingly connect the next population’s optimum values. In case of uneven the HUIs distribution are uneven, previous population’s optimum values are to be target search meant by some missing results inside certain number of iterations. Thus, Bio-HUIF-Improved Particle Swarm Optimization (Bio-HUIF-IPSO) is proposed for discovering HUI for target source of the sequence population with the ratio of its utility to the total utilities of all discovered HUIs accordingly. Discovery process in HUI is accelerated by bitmap database representation strategies. Improved Particle Swarm Optimization(IPSO) algorithm are validated on real datasets with wide-ranging experiments.