Performance Evaluation of Decision Tree Classifiers on Medical Datasets

In data mining, classification is one of the significant techniques with applications in fraud detection, Artificial intelligence, Medical Diagnosis and many other fields. Classification of objects based on their features into predefined categories is a widely studied problem. Decision trees are very much useful to diagnose a patient problem by the physicians. Decision tree classifiers are used extensively for diagnosis of breast tumour in ultrasonic images, ovarian cancer and heart sound diagnosis. In this paper, performance of decision tree induction classifiers on various medical data sets in terms of accuracy and time complexity are analysed.