A novel gene selection method for multi-catalog cancer data classification

Abstract

In this paper, a novel gene selection method which was merging the relevance score (BW ratio) and the Flexible Neural Tree (FNT) together was proposed for the multi-class cancer data classification. Firstly, the BW ratio method was adopted to select some informative genes, and then the FNT method was used to extract more characteristic genes from the gene subsets. FNT is a tree-structured neural network with input variables selection, over-layer connections and different activation functions for different nodes. Based on the pre-defined instruction/operator sets, a flexible neural tree model can be created and evolved. The FNT structure is developed by using probabilistic incremental program evolution (PIPE) algorithm, and the free parameters embedded in neural trees are optimized by particle swarm optimization (PSO) algorithm. Experiment on two well-known cancer datasets shows that the proposed method achieved better results compared with other methods. © 2012 Springer-Verlag.