A Classification of Microarray Gene Expression Data Using Hybrid Soft Computing Approach

In this paper, an efficient technique is proposed for the precise classification of microarray genes from the microarray gene expression dataset. The proposed classification technique performs the classification process with the aid of three phases namely, dimensionality reduction, feature selection and gene classification. Initially, the proposed technique reduces the dimensionality by utilizing Genetic Algorithm (GA). The main objective of dimensionality reduction is to select the optimal number of genes from the microarray gene expression dataset. Next, in the feature selection process the features are extracted from the column gene values. Here, probability of GA-indexed gene and new statistical features are selected for each column gene values and these selected features are given to the Feed Forward Back propagation Neural Network (FFBNN). The FFBNN network is trained using the selected features and then this well trained FFBNN network performance is tested with the column gene values. The FFBNN network classifies the microarray gene values into their corresponding cancer class types. The performance of the classification technique is evaluated by the performance measures such as accuracy, specificity and sensitivity.