A Successive Feature Selection Algorithm for Gene Ranking

Abstract: Identification and classification of cancer for the gene is most vital. The importance of the each gene is to be found by the gene raking measurement. Modified Successive Feature Selection is used for gene ranking in this paper. Then the Support Vector Machine classifier is trained with that dataset. Genes are collected from the dataset. Many of the feature selection algorithms produced fault for their ranked gene performance. To prevent this, proposed method produces the better accuracy by producing a feature selection algorithm in gene expression data analysis of sample classifications. That the proposed method selects the gene and divides the genes into subset, from the features, gene ranks are selected. From the Lymphoma and Leukemia dataset genes are selected. The proposed method shows promising classification accuracy for the entire test data sets.

Keywords: Successive Feature Selection, Modified Successive Feature Selection, Support Vector Machine, Lymphoma dataset, Leukemia dataset.