Soft clustering and support vector machine based technique for the classification of abnormalities in digital mammograms

This paper presents a novel technique which is the amalgamation of a clustering mechanism and a Support Vector Machine classifier. The technique is called Soft Clustering based Support Vector Machine and is designed to provide a fast converging network with good generalization ability leading to an appropriate classification as a benign or malignant class for the classification of suspicious areas in digital mammograms. The proposed technique has been evaluated on a benchmark database. The experimental results and analysis of results are included in this paper.