Identification of Salient Patterns for Classification of Gene Expression Data

Identification of salient patterns for the classification of gene expression profiles is a useful step in examining the biological significance and correlation of genes with disease states. We propose a clustering-based approach in which feature selection is first carried out to identify influential genes and then salient patterns are determined to characterize each of the different classes. The proposed method has been tested with the complicated colon tumor data and the experimental results are evaluated in comparison with the published ones.