Feature selection in cancer microarray data using multi-objective genetic algorithm combined with correlation coefficient

摘要

Cancer microarray data consists of an enormously huge number of features, most of which are irrelevant for classifying microarray gene expression patterns. Selection of a minimal subset of features for classification is a challenging task. In this study, in order to achieve faster convergence, an initial dimensionality reduction is performed based on the correlation coefficients of features. Multi-objective Genetic Algorithm is then applied to select a minimal set of non-redundant features. The effectiveness of this approach is demonstrated on three well-known cancer datasets.