Feature selection and mass classification using particle swarm optimization and support vector machine

Abstract

This paper proposes an effective technique to classify regions of interests (ROIs) of digitized mammograms into mass and normal breast tissue regions by using particle swarm optimization (PSO) based feature selection and Support Vector Machine (SVM). Twenty-three texture features were derived from the gray level co-occurrence matrix (GLCM) and gray level histogram of each ROI. PSO is used to search for the gamma and C parameters of SVM with RBF kernel which will give the best classification accuracy, using all the 23 features. Using the parameters of SVM found by PSO, PSO based feature selection is used to determine the significant features. Experimental results show that the proposed PSO based feature selection technique can find the significant features that can improve the classification accuracy of SVM. The proposed classification approach using PSO and SVM has better specificity and sensitivity when compared to other mass classification techniques.