Predicting disease risks from highly imbalanced data using random forest

Table 8 RF, SVM, bagging and boostingperformance without sub-sampling in terms of AUC on eight disease categories

Disease	RF	SVMSVM	Bagging	Boosting
Breast cancer	0.8793	0.5	0.5085	0.836
Diabetes no complication	0.8567	0.5	0.4749	0.8175
Diabetes with/complication	0.9084	0.648	0.4985	0.8278
Hypertension	0.8893	0.6908	0.4886	0.8515
Coronary Atherosclerosis	0.9193	0.6601	0.4945	0.8608
Peripheral Atherosclerosis	0.8872	0.5	0.4925	0.8279
Other Circulatory Diseases	0.7389	0.5	0.4829	0.6851
Osteoporosis	0.7968	0.5	0.4931	0.8561

ISSN: 1472-6947