Constructing a Predictive Model for Detection of Breast Cancer
Cancer is a big issue all over the world. It is a disease, which is incurable in many cases and has affected the lives of many and will continue to affect the lives of many more. Breast cancer is the second leading cause of cancer deaths in women now a day and has become the most common cancer among women both in the developed and the developing world. Early detection is the most effective way to decrease breast cancer deaths. But early detection needs an accurate and reliable diagnosis procedure that allows doctors to differentiate benign breast tumors from malignant ones without going for surgical biopsy. Hence, construct a predictive model using data mining techniques to identify hidden knowledge and develop a prototype interface for breast cancer that support health professional in their diagnosis decisions and treatment planning measures. For this study, a six-step hybrid knowledge discovery process model is followed, due to the nature of the problem and attributes in the dataset. The classification technique such as, J48 decision tree, Naive Bayes and PART rule induction used to build the models. Performance of the models is compared using accuracy, TPR, TNR, and the area under the ROC curve. J48 decision tree registers better performance with 94.82 % accuracy.
Keywords - Benign, Breast Cancer, classification, Data mining, Knowledge Discovery in Databases, Malignant, WEKA