Performance Comparison of Artificial Neural Network techniques for Classification and Prediction of Breast Cancer

Abstract

Data Mining is used to retrieve the information from the huge amount of data. Data mining is important field from all the fields for providing accurate prediction of diseases and deeper study of medical data. Data mining application involves analysing Breast cancer, which are the most dangerous disease cause of cancer deaths in women worldwide. Artificial Neural Network is used here to predict the breast cancer by analysing the nine technical indicators like clump thickness, uniformity of cell size, uniformity of cell shape, marginal adhesion, single epithelial cell size, bare nuclei, bland chromatin, normal nuclei and mitoses[3]. This study compares the four models namely Pattern Recognition Networks, Feed Forward Back Propagation Networks, Feed Forward Networks with no feedback, and Radial Basis Function Network to predict the breast cancer. The performance of all these models are examined from accuracy measures like Mean Square Error, Mean Absolute Error, Sum Square Error and Root Mean Square Error. From the examined result, the performance of Radial Basis Function network was considered as a best network for predicting the breast cancer. Keywords: Data mining, Breast Cancer, \xa0Feed Forward Back Propagation networks, Feed Forward networks, Pattern Recognition networks Radial Basis Function, Technical Indicators .