Microarray Gene Ranking Technique Based on Modified Successive Feature Selection Algorithm

Abstract

The most important activities in genetic algorithm are Cancer classification and Identification.  The significance of each gene can be measured by ranking the genes. In this paper, Modified Successive Feature Selection is used for ranking the genes and the classifier is trained with the genes  collected from the dataset. Many of the existing feature selection algorithms produce errors while ranking the gene performance. A new method with better accuracy is proposed to prevent it  by using a modified feature selection algorithm in the data analysis of gene expression data.  The proposed method selects the genes and divides them from set into subset and, from the ranked features the genes are selected. Lymphoma and Leukemia dataset are used  as a sample dataset to select the genes. The proposed method shows a promising classification accuracy for the entire test data sets
Copyright © 2014 Praise Worthy Prize - All rights reserved.