Classification of Breast Cancer Masses using Non-Linear Quadratic Support Vector Machine and Comparison with Self-Organizing Neural Network

Abstract

Breast cancer is the second most common cancer after lung cancer and the fifth leading cause of death in women. In less developed countries, breast cancer is the most important cause of death. In this disease, the cells of the breast tissue change and divide into multiple cells and cause a lump. If breast cancer is in the early stages, treatment is possible. There are many treatment methods such as surgery to remove the defective area, drug therapy, radiation therapy, chemotherapy, hormone therapy, and immunotherapy. These treatments have the potential to save lives when administered in the early stages. From the above explanations, it can be seen that early detection of breast cancer is very important and in this research, an attempt has been made to identify suspected cancer data with the quadratic support vector machine method and based on the features extracted from valid and numerous MRI images. Let's classify so that the process of diagnosing the disease in the early stages is easier and faster. The results showed that 356 out of 357 malignant data and 202 out of 211 benign data were correctly classified. The classification accuracy of malignant data was 99.7% and the classification accuracy of benign data was 97.5%, and finally the overall classification accuracy was 98.2%, which indicates the optimal performance of this method in breast cancer data classification.

Keywords

• breast cancer
• wisconsin data
• support vector machine
• quadratic kernel.