Table 2.

The highest values for accuracy, AUC, F-measure, precision, recall, sensitivity, and specificity for predicting responders vs. non-responders (A) and responders vs. relapsers (B) groups

A
	Subtype 1a (Responders vs. Non-Responders)				Subtype 1b (Responders vs. Non-Responders)
	Bayes	Neural Networks	SVM	Decision Trees	Bayes	Neural Networks	SVM	Decision Trees
Database	Chi Squared	SVM	Relief	PCA	SVM	SVM	Relief	Gini Index
Algorithm	Naive Bayes (Kernel)	AutoMLp	SVM	DT Parallel Gini Index	Naive Bayes (Kernel)	AutoMLp	SVM	DT Random Forest Info Gain
Accuracy	74.17%	76.67%	74.17%	69.17%	89.17%	85.00%	75.00%	80.00%
AUC	0.84	0.68	0.75	0.59	0.94	0.94	0.84	0.83
AUC (optimistic)	0.84	0.68	0.75	0.83	0.94	0.94	0.84	0.85
AUC (pessimistic)	0.84	0.68	0.75	0.58	0.94	0.94	0.84	0.80
F-Measure	0.78	0.82	0.80	0.73	0.92	0.87	0.80	0.86
Precision	0.84	0.82	0.80	0.80	0.93	0.94	0.81	0.87
Recall	0.73	0.82	0.88	0.73	0.93	0.83	0.83	0.90
Sensitivity	0.73	0.82	0.88	0.73	0.93	0.83	0.83	0.90
Specificity	0.85	0.75	0.60	0.65	0.85	0.80	0.50	0.65
B
	Subtype 1a (Responders vs. Relapsers)				Subtype 1b (Responders vs. Relapsers)
	Bayes	Neural Networks	SVM	Decision Trees	Bayes	Neural Networks	SVM	Decision Trees
Database	Chi Squared	SVM	Relief	PCA	SVM	SVM	Relief	Gini Index
Algorithm	Naive Bayes (Kernel)	AutoMLp	SVM	DT Parallel Gini Index	Naive Bayes (Kernel)	AutoMLp	SVM	DT Random Forest Info Gain
Accuracy	82.50%	79.17%	82.50%	81.67%	78.33%	78.33%	84.17%	81.67%
AUC	0.89	0.79	0.82	0.61		0.00		0.66
AUC (optimistic)	0.89	0.79	0.82	0.91		0.85		0.85
AUC (pessimistic)	0.89	0.79	0.82	0.74		0.15		0.47
F-Measure	0.84	0.86	0.86	0.84	0.87	0.87	0.91	0.89
Precision	0.92	0.83	0.90	0.90	0.78	0.78	0.84	0.82
Recall	0.82	0.92	0.87	0.80	1.00	1.00	1.00	1.00
Sensitivity	0.82	0.92	0.87	0.80	1.00	1.00	1.00	1.00
Specificity	0.85	0.55	0.75	0.85	0.00	0.00	0.29	0.14