TABLE 4.

Accuracy of 101 times classification with three different combinations as input features of decision tree, and confusion matrix of the classification represented by the median of the 101 accuracy values by subject class (FESZ, UHE and HC).

	Accuracy	FESZ	UHR	HC	Total
					P/T	%
DEM, MCCB	53.90 ± 9.78	3/5	2/5	3/4	8/14	57.14
DEM, MCCB, ERP	64.29 ± 9.03∗	3/5	4/5	2/4	9/14	64.29
DEM, MCCB, ERP, NET	79.22 ± 10.81∗	4/5	4/5	4/4	12/14	85.71

^∗Indicates a significant difference compared to the previous group, p < 0.05. The denominator (T) represents the number of samples in the test set, and the numerator (P) represents the number of samples that are predicted correctly. DEM, demographic data; NET, brain network parameters; MCCB, MCCB indexes; ERP, P50 performance.