Breast Cancer Factors Detection Using Classification Algorithms

Breast cancer is a major concern among women. It is important that health care organizations are equipped with reliable software and statistical tools to analyze the vast amount of available data that can aid in the early detection of breast cancer. This paper demonstrates how the Knowledge Discovery in Databases (KDD) Methodology can be enlisted to filter such data in order to select the major attributes identified as significant factors in the diagnosis and treatment of breast cancer.: Following this methodology, five classifiers will be tested out by using the WEKA software and the accuracy rate of each classifier applied to the dataset of breast cancer will be observed. These are Naive Bayes, random forest, decision tree, ZeroR and OneR. The accuracy rate of each as applied to the dataset of breast cancer markers was observed. Based on these trials, the Naive Bayes classifier has been selected as the best data mining method.