Predicting disease risks from highly imbalanced data using random forest

Table 5 RF, SVM, bagging and boosting performance in terms of AUC on eight disease categories

Disease	RF	SVM	Bagging	Boosting
Breast cancer	0.9123	0.9063	0.905	0.8886
Diabetes no complication	0.8791	0.8417	0.8568	0.8607
Diabetes with/complication	0.94317	0.9239	0.9294	0.9327
Hypertension	0.9003	0.8592	0.8719	0.8842
Coronary Atherosclerosis	0.9199	0.8973	0.887	0.9026
Peripheral Atherosclerosis	0.9095	0.8972	0.8967	0.9003
Other Circulatory Diseases	0.7899	0.7591	0.7669	0.7683
Osteoporosis	0.87	0.867	0.8659	0.8635

ISSN: 1472-6947