PERFORMANCE ANALYSIS OF LOGISTIC REGRESSION AND KERNEL LOGISTIC REGRESSION FOR BREAST CANCER CLASSIFICATION

Breast Cancer is one of the most commonly occurring cancers among women. There are various types of breast cancer that differ in their capability of spreading metastasize to other tissues of the body. The exact cause of breast cancer is not fully understood but a number of risk factors have been identified. The signs and symptoms of breast cancer include a lump in the armpit or breast, blood discharges in the nipple, inverted nipple, dimpling of the breast skin, pain in the breast, sore nipple, swollen lymph node in the neck, changes in the size and shape of the breast. Generally, breast cancer is diagnosed during a physical examination of the breast, ultrasound testing, mammography and biopsy. Breast Cancer treatment depends on the type of cancer and its specific stage and involves radiation, surgery or chemotherapy. In this paper, Logistic Regression and Kernel Logistic Regression are used as post classifiers to classify the breast cancer. Results show that an average classification accuracy of 97.75% is obtained when Logistic Regression is used and an average classification accuracy of 94.28% is obtained when Kernel Logistic Regression is utilized.