An Ensemble Learning Method on Mammogram Data for Breast Cancer Prediction—Comparative Study

Abstract

Cancer remains a serious problem for people nowadays because of its largest death rate in the world. The biggest rate is full of the absence of adequate early detection of cancer. Breast cancer carcinoma is one of the main killer illness among ladies in the world, and the commonly diagnosed non-skin cancer in women. Early detection of this deadly disease can considerably increase the possibility of successful treatment. The traditional way of cancer diagnosis primarily depends on the doctor's or technician's experience to identify the abnormalities in our human body. Bosom malignancy happens once the cell tissues of the breast become irregular and wildly divided. By investigating, the mammogram information can help doctors in recognizing disease cases at a whole lot sooner stage and accordingly can altogether improve the capability of patient treatment. It is as yet a test to anticipate the forecast of disease patients, due to its high heterogeneity and multifaceted nature. This study focuses on the forecast of breast cancer from mammogram pictures and assesses the precision of various CNN models.