Table 2.

Summary of the relative performance of each model in the percent of testing subjects classified correctly indicated by mean (standard deviation)—max

	Overall	Healthy	GoF	LoF
Decision tree
Accuracy	62.05 (1.54)–65.38	60.0 (6.92)–69.23	70.77 (8.8)–80.77	55.38 (9.3)–73.08
True Positive Rate		0.6 (0.07)–0.69	0.71 (0.09)–0.81	0.56 (0.09)–0.73
False Positive Rate		0.23 (0.05)–0.31	0.17 (0.05)–0.25	0.17 (0.05)–0.25
F1		0.58 (0.04)–0.64	0.69 (0.04)–0.76	0.58 (0.06)–0.7
Gaussian Naive Bayes
Accuracy	65.98 (2.07)–70.51	83.19 (6.1)–100.0	68.88 (5.65)–80.77	45.86 (6.85)–61.54
True Positive Rate		0.83 (0.06)–1.0	0.69 (0.06)–0.81	0.46 (0.07)–0.62
False Positive Rate		0.33 (0.06)–0.48	0.07 (0.04)–0.17	0.11 (0.04)–0.23
F1		0.67 (0.03)–0.75	0.75 (0.04)–0.83	0.55 (0.05)–0.62
Neural network
Accuracy	53.2 (1.5)–57.69	53.85 (7.45)–65.38	61.78 (7.38)–73.08	43.99 (9.42)–57.69
True Positive Rate		0.54 (0.08)–0.65	0.62 (0.07)–0.73	0.44 (0.09)–0.58
False Positive Rate		0.24 (0.05)–0.33	0.19 (0.06)–0.37	0.27 (0.06)–0.37
F1		0.53 (0.04)–0.59	0.62 (0.04)–0.67	0.44 (0.06)–0.53
Support vector machine
Accuracy	60.34 (2.66)–65.38	72.46 (10.18)–92.31	60.18 (7.3)–76.92	48.37 (12.3)–73.08
True Positive Rate		0.72 (0.1)–0.92	0.6 (0.07)–0.77	0.48 (0.12)–0.73
False Positive Rate		0.31 (0.09)–0.58	0.11 (0.04)–0.19	0.18 (0.07)–0.33
F1		0.62 (0.04)–0.71	0.66 (0.06)–0.77	0.52 (0.08)–0.67
Gradient boosting decision tree
Accuracy	62.55 (1.23)–66.67	59.16 (7.89)–76.92	66.32 (6.59)–80.77	62.15 (7.28)–76.92
True Positive Rate		0.59 (0.08)–0.77	0.66 (0.06)–0.81	0.62 (0.07)–0.77
False Positive Rate		0.22 (0.05)–0.35	0.14 (0.04)–0.25	0.2 (0.05)–0.31
F1		0.58 (0.04)–0.68	0.68 (0.04)–0.79	0.61 (0.05)–0.73