Metaheuristics for Frequent and High-Utility Itemset Mining

Abstract

Metaheuristics are often used to solve combinatorial problems. They can be viewed as general purpose problem-solving approaches based on stochastic methods, which explore very large search spaces to find near-optimal solutions in a reasonable time. Some metaheuristics are inspired by biological and physical phenomenons. During the last two decades, two population-based methods named evolutionary algorithms and swarm intelligence have shown high efficiency compared to many other metaheuristics. Frequent Itemset Mining (FIM) and High Utility Itemset Mining (HUIM) are the process of extracting useful frequent and high utility itemsets from a given transactional database. Solving FIM and HUIM problems can be very time consuming, especially when dealing with large-scale data. To deal with this issue, different metaheuristic-based methods were developed. In this chapter, we study the application of metaheuristics to FIM and HUIM. Several metaheuristics have been presented, based on evolutionary or swarm intelligence algorithms, such as genetic algorithms, particle swarm optimization, ant colony optimization and bee swarm optimization.