Machine Learning based System for Prediction of Breast Cancer Severity

Breast cancer is one of the most common diseases and the leading cause of death to mostly females all over the world. Early detection can provide higher treatment efficiency and better healing chances. Even though mammography screening is handy in diagnosing breast cancer at an early stage, Computer-Aided Diagnosis (CAD) systems can help to reduce the cancer death-rate. Radiologists, physicians, and doctors, in general, make use of these CAD systems to diagnose, detect, analyze and make decisions whether the patient is benign or malignant. The present paper presents some data mining techniques used in the diagnosis of cancer such as Artificial Neuron Network (ANN), K-Nearest Neighbors (KNN), Binary Support Vector Machine (Binary SVM), and Decision Tree (DT). Within this framework, the database utilized is the Mammographic Mass dataset. This database contains data of probabilistic breast cancer patients and the advanced results by experts in the field. The paper adopts a confusion matrix for binary prediction as a method of data analysis. The present paper provides a comparison between the different Computer-Aided diagnosis systems techniques regarding accuracy, specificity, and sensitivity amidst many other criteria to find the most accurate alternative among ANN, KNN, Binary SVM, and DT.