Meta-heuristic search based gene selection and classification of microarray data

Microarray dataset often contains huge number of data, only a fraction of which comprise significant differentially expressed genes. In this article, t-test is applied to identify the precise and interesting genes which are responsible for cause of cancer. After precise identification, a new embedded approach is proposed where in a genetic algorithm (GA) is combined with Artificial neural network (ANN) along with t-test. This ANN-based GA has the both characteristic, to identify a small subset of informative gene from the initial data set in order to obtain high accuracy. Further, this paper presents a comparative analysis on the obtained classification accuracy by the proposed approach and other existing models available in the literature. From the obtained results, it is apparent that proposed approach is able to achieve high accuracy with a small number of selected genes.