Type-II Diabetes detection using Decision-tree based Ensemble of Classifiers

Type 2 diabetes also known as diabetes mellitus is a condition that affects the way body processes blood sugar. Researchers are working at improving the detection accuracy of machine learning based model for detection of type 2 diabetes. Dataset available to train the classifier imposes severe challenges in this research due to observed missing values, unbalanced data. This paper examines the effectiveness of Decision Tree based Ensemble classifiers for improving the detection accuracy of type 2 diabetes by tuning them using assorted combinations of parameter values. The results of early studies and Decision Tree based Ensemble Classifiers is compared and it can be seen that performance of the later is better. Experimental results suggest that Random Forest and XGBoost classifier gives better performance as compared to other Decision Tree-based ensemble classifiers.