Tackling the biclustering problem with cooperative coevolutionary algorithms

The biclustering problem consists in simultaneously clustering rows and columns of a data matrix. The aim of this paper is to empirically assess the performance of cooperative coevolution as an alternative approach for coping with the task of discovering good and sizeable biclusters. For this purpose, two cooperative coevolutionary algorithms, one configured with genetic algorithms (GAs) and another configured with particle swarm optimization (PSO), have been investigated through experiments conducted over two real-world problems. The results achieved reveal that, when compared with simple GA, standard PSO, and the PSO-GA hybrid algorithms, the coevolutionary models (especially the ones configured with PSO) usually prevail in terms of discovering larger coherent biclusters, but lag behind in terms of computational efficiency.