Investigations on feature selection techniques in microarray mirna expression data for cancer classification

Abstract

Cancer is a complex disease where molecular mechanism and signalling pathways are elusive and intangible. A dynamic and systematic approach is required to integrate and interrogate diverse biological information for the prognosis and therapeutic interventions. It is essential to understand the gene interactions in signalling pathways, molecular networks and functional attributes to unravel the biological behaviour of tumours. The newlinemicroarray data experiment contains large number of features and small number of samples. The high dimensionality of the DNA microarray data and smaller sample size than gene size becomes problem, when it is employed for classification of cancer. Cancer Classification from microarray expression profiles is a newlinechallenging task due to its high dimensionality in the field of biomedicine and bioinformatics. So the feature selection which provides highly informative and discriminative feature set is an essential task in cancer classification. The microarray classification is a two step process. The first step is to select a subset of significant and relevant genes from the set of genes and the secondstep is to develop a classification model that can produce accurate prediction for new data. A true and accurate classification is important for the successful diagnosis and treatment of cancer newline newline