Pattern Ensemble Learning Method for Clustering Ensemble using Incremental Genetic-Based Algorithm

Abstract

The clustering ensemble has emerged as a prominent method for improving clustering accuracy of unsupervised classification. It combines multiple partitions generated by different clustering algorithms into a single clustering solution. Genetic algorithms are known as methods with high ability to solve optimization problems including clustering. To date, significant progress has been contributed to find consensus clustering that will yield better results than existing clustering. This paper has proposed an Incremental Genetic-Based Algorithm for Clustering Ensemble (IGCE) to perform the search task, but has replaced its traditional crossover operator with a Pattern Ensemble Learning Method (PEL). Therefore, IGCE-PEL is capable to avoid the problems of clustering invalidity and context insensitivity from the traditional crossover operator of genetic algorithms. IGCEs have been evaluated on twelve benchmark datasets based on different recombination operators used. The experimental results have demonstrated that IGCE using PEL is able to achieve better clustering accuracy when compared with several other existing genetic-based clustering ensemble algorithms.