Feature selection with improved binary artificial bee colony algorithm for microarray data

Abstract

In the areas of clinical diagnosis, gene expressions are known to have latent qualities as they denote the state of cells in molecular rankings. But the sample sizes are relatively small compared to the number of genes. Hence, the need to develop an efficient gene selection algorithm is appropriate to enhance predictive accuracy and as well prevent unfathomable conditions from the extensive quantity of genes. This article proposes an improved binary artificial bee colony algorithm (BABC) based on chaotic catfish effect for feature selection. Chaotic effect was added to the initialisation procedure of BABC, and further introduced chaotic catfish-bee for new nectar exploration, which can thus improve the BABC algorithm by preventing bees from getting trapped in a local optimum. The experiment shows that this new method indicated an elaborate feature simplification which achieved a very precise and significant accuracy of nine among the 11 datasets compared with other methods.