Breast Cancer Classification and Biomarker Discovery on Microarray Data Using Genetic Algorithms and Bayesian Classifier

Abstract

In this paper, we aim at using genetic algorithms (GAs) for gene selection and propose Bayes’ theorem as a discriminant function to classify breast cancers for biomarker discovery. The GA is used to search all possible gene subsets from microarray feature dimensions. To evaluate the given feature subsets, a Bayesian discriminant function is developed to produce a classifier to measure the fitness of each gene subset. And these values will be stored to provide feedback for the evolution process of GA to find the increasing fit of chromosomes in the next generation. Consequently, the experimental results show that our method is effective to discriminate breast cancer subtypes and find many potential biomarkers to help cancer diagnosis.