Empirical study on the performance of the classifiers based on various criteria using ROC curve in medical health care

Classification is one of the most efficient data mining techniques in Machine Learning. In classification, Decision trees can handle high dimensional data. But, decision trees yield poor performance in medical health care. So, In this paper, we investigate the use of Receiver Operating Characteristic (ROC) curve for the evaluation of machine learning algorithms. In particular, we investigate the use of the area under the ROC curve (AUC) as a measure of classifier performance. AUC help to determine decision tree characteristics, such as node selection, misclassification error, cost parameter and stopping criteria. In this paper, we empirically evaluate the performance of ROC and 2 decision tree algorithms on the cancer dataset taken from the UCI ML Repository.