An Entropy-based gene selection method for cancer classification using microarray data

Background  

Accurate diagnosis of cancer subtypes remains a challenging problem. Building classifiers based on gene expression data is a promising approach; yet the selection of non-redundant but relevant genes is difficult.