A Hybrid Multi Objective Particle Swarm Optimization Method to Discover Biclusters in Microarray Data

TL;DR

This paper introduces a hybrid multi-objective particle swarm optimization method for biclustering in microarray data, improving overlap and coverage in gene expression analysis.

Contribution

It presents a novel hybrid algorithm combining multi-objective PSO for biclustering, addressing conflicting objectives and enhancing gene data coverage.

Findings

Significant reduction in bicluster overlap.

Improved coverage of gene expression matrix elements.

Enhanced biclustering performance on benchmark datasets.

Abstract

In recent years, with the development of microarray technique, discovery of useful knowledge from microarray data has become very important. Biclustering is a very useful data mining technique for discovering genes which have similar behavior. In microarray data, several objectives have to be optimized simultaneously and often these objectives are in conflict with each other. A Multi Objective model is capable of solving such problems. Our method proposes a Hybrid algorithm which is based on the Multi Objective Particle Swarm Optimization for discovering biclusters in gene expression data. In our method, we will consider a low level of overlapping amongst the biclusters and try to cover all elements of the gene expression matrix. Experimental results in the bench mark database show a significant improvement in both overlap among biclusters and coverage of elements in the gene expression matrix.