Minimal Feature Selection using Statistical Techniques

Cancer Classification based on microarray gene expression is an important problem. In this work, we use a ANOVA bsed feature selection method we choose some important genes form thousands of genes, after that, we classify the microarray data sets with a fuzzy neural network (FNN) that we proposed earlier. This FNN combines important features of initial fuzzy model selforganization, parameter optimization and rule-base simplification. We applied this FNN to well-known gene expression data sets i.e., the Lymphoma data set with 3 sub-types. Our result in the lymphoma data set show that the FNN can obtain 100% accuracy with a much smaller number of genes in comparison with previously published Our simple yet very effective method involves two steps. In the first step, we choose some important genes using an Analysis of Variance (ANOVA) ranking scheme. In the second step, we test the classification capability of all simple combinations of those important genes using a good classifier. Our approach obtained very high accuracy with only two genes.