Informative gene selection for microarray classification via adaptive elastic net with conditional mutual information

Abstract

• AEN-CMI is a hybrid and penalization method that is built on the popular AEN method. • AEN-CMI incorporates the conditional mutual information, into the adaptive weight estimation process. • AEN-CMI encourages a grouping effect and could be solved by a widely available algorithm: PCD. • AEN-CMI performs better on colon cancer and leukemia cancer datasets than SVM, classic elastic net , adaptive Lasso and adaptive elastic net. • AEN-CMI obtains the highest classification accuracy by using the least number of genes. Due to the advantage of achieving a better performance under weak regularization , elastic net has attracted wide attention in statistics, machine learning, bioinformatics , and other fields. In particular, a variation of the elastic net, adaptive elastic net (AEN), integrates the adaptive grouping effect. In this paper, we aim to develop a new algorithm: Adaptive Elastic Net with Conditional Mutual Information (AEN-CMI) that further improves AEN by incorporating conditional mutual information into the gene selection process. We apply this new algorithm to screen significant genes for two kinds of cancers: colon cancer and leukemia. Compared with other algorithms including Support Vector Machine , Classic Elastic Net, Adaptive Lasso and Adaptive Elastic Net, the proposed algorithm, AEN-CMI, obtains the best classification performance using the least number of genes.