Tissue classification using gene expression data and artificial neural network ensembles

An important challenge in the use of large-scale gene expression data for biological classification occurs when the number of genes far exceeds the number of samples. This situation will make the classification results are unstable. Thus, a tissue classification method using artificial neural network ensembles was proposed. In this method, a feature preselection method is presented to identify significant genes highly correlated with tissue types. Then pseudo data sets for training the component neural network of ensembles were generated by bagging. The predictions of those individual networks were combined by simple averaging method. Some data experiments have shown that this classification method yields competitive results on several publicly available datasets.