Cross-Entropy Assisted Optimization Technique for High Utility Itemset Mining from the Transactional Database

Abstract

High Utility Itemset Mining (HUIM) is the process of discovering profitable itemsets in a transactional database with a high utility or profit range. This technique is mainly used to unearth the prominent patterns and insights from large volumes of data. The main goal of HUIM is to utilize optimization techniques that improve the efficiency of mining operations. This research article introduces a Genetic Algorithm (GA) integrated Bacterial Foraging Algorithm (BFA) based on the HUIM technique, which is efficient due to its exploitation and exploration ability. This incorporation intends to address the issue of early convergence in itemset mining, and Cross Entropy (CE) act as a genetic operator, which enhances the efficiency of the mining operation. The technique effectively perceives High Utility Itemsets (HUIs) within a significantly reduced duration of 16723 ms, accomplishing optimal outcomes. The strategy effectively reduces the likelihood of premature convergence and showcases exceptional efficiency, as seen by the successful retrieval of 3432 optimized itemsets. The suggested hybrid GA-BFA approach effectively addresses problems in itemset mining by achieving high-quality results in a small amount of time. The implementational outcomes of the proposed Hybridized Cross Entropy Assisted Genetic algorithm with Bacterial Foraging Algorithm (HCEG-BFA) are compared with existing state-of-art techniques where the HCEG-BFA outperforms existing state-of-art techniques.