Feature selection and classification for gene expression data analysis

Abstract

In recent years rapid developments in genomics and other biological research have generated a huge amount of biological data This data requires sophisticated computational analysis in order to interpret the data and to draw conclusions One of the most active areas of research in bioinformatics is the use of data mining techniques to solve biological problems Examples of biological analysis include protein structure prediction statistical modeling of protein protein interactions analysis of mutations and cancer classification based on microarray data This research employs data mining feature selection methods to discover the differentially expressed genes from gene expression microarray datasets for use in classification Microarrays are used to measure the expression levels of large numbers of genes simultaneously most of them are uninformative and redundant The sample sizes are very low relative to the number of genes which reduces the diagnosis accuracy of statistical models Thereby selecting highly significant genes from gene expression data improves the classification performance This thesis aims to identify a subset of genes that are relevant for model construction among a huge number of genes In this thesis two pre processing methods and two filter based feature selection techniques are proposed newline