Homogenous Ensembles on Data Mining Techniques for Breast Cancer Diagnosis

Abstract

Breast cancer is a disease usually found in women which poses serious health challenges and can be fatal if not diagnosed quickly and treated immediately. Techniques of data mining have been defined to play a significant role in the diagnosis of numerous diseases in which breast cancer diagnosis is a good example. This paper employed homogenous ensemble on methods of data mining for breast cancer diagnosis. Three data mining classification algorithms: k-nearest neighbour, Decision Tree (C4.5) and Support Vector Machines (SVM) with their homogenous ensembles of Bagging and Boosting were applied. The experimental result showed that support vector machines possess the highest classification accuracy, homogenous ensembles of bagging and boosting does not affect classification accuracy greatly as they either slightly increase or reduce the accuracy of classification while increasing the time it takes for the algorithm to build its model.