Breast Tumor Classification Using SVM

Abstract

Although there are several techniques that have been used to differentiate between benign and malignant breast tumor lately, support vector machines (SVMs) have been distinguished as one of the common method of classification for many fields such as medical diagnostic, that it offers many advantages with respect to previously proposed methods such as ANNs. One of them is that SVM provide a higher accuracy, another advantage that SVM reduces the computational cost, and it is already showed good result in this work. In this paper, a Support Vector Machine for differentiation Breast tumor was presented to recognize malignant or benign in mammograms. This work used 569 cases and they were classified into two groups: malignant (+1) or benign (-1), then randomly selected some of these samples for training model while others were used for test. The ratios were 84.4.0% of accepted false, 947142% of refused false. These results indicate how much this method is successful.

Keywords

• SVM
• Breast tumor
• Classification
• Benign
• Malignant