Table 3. Performance of 25 classification algorithms.

Model ID	Model Name	8ROIs_accuracy	8ROIs_AUC	Sensitivity	Specificity
1	Fine Tree	85.1	0.88 (0.13, 0.79)	79	88
2	Medium Tree	85.1	0.88 (0.13, 0.79)	79	88
3	Coarse Tree	80.6	0.86 (0.19, 0.79)	79	81
4	Linear Discriminant	78.4	0.81 (0.09, 0.47)	47	91
5	Quadratic Discriminant	76.9	0.72 (0.13, 0.5)	50	88
6	Logistic Regression	76.9	0.8 (0.09, 0.42)	42	91
7	Gaussian Naïve Bayes	79.1	0.85 (0.18, 0.71)	71	82
8	Kernel Naïve Bayes	79.1	0.85 (0.18, 0.71)	71	82
9	Linear SVM	82.1	0.82 (0.09, 0.61)	61	91
10	Quadratic SVM	79.1	0.73 (0.09, 0.50)	50	91
11	Cubic SVM	73.9	0.74 (0.17, 0.50)	50	83
12	Fine Gaussian SVM	71.6	0.73 (0.00, 0.00)	0	100
13	Medium Gaussian SVM	80.6	0.80 (0.09, 0.55)	55	91
14	Coarse Gaussian SVM	80.6	0.85 (0.08, 0.53)	53	92
15	Fine KNN	76.9	0.70 (0.14, 0.53)	53	86
16	Medium KNN	79.1	0.81 (0.15, 0.63)	63	85
17	Coarse KNN	71.6	0.78 (0.00, 0.00)	0	100
18	Cosine KNN	76.9	0.82 (0.22, 0.74)	74	78
19	Cubic KNN	81.3	0.81 (0.13, 0.66)	66	88
20	Weighted KNN	77.6	0.82 (0.13, 0.53)	53	88
21	Boosted Trees	71.6	n/a	0	100
22	Bagged Trees	75.4	0.78 (0.16, 0.53)	53	84
23	Subspace Discriminant	80.6	0.82 (0.08, 0.53)	53	92
24	Subspace KNN	76.9	0.80 (0.15, 0.55)	55	85
25	RUSBoosted Trees	77.6	0.83 (0.20, 0.71)	71	80

ROIs: regions of interest. AUC: area under curve.