Table 5.

Summary results for all two-class classification using leave-one-out cross-validation (LOOCV) for 10-sec time-window size.

Classification Tasks	10-sTime Window Size
	Acc. (%)			Sens. (%)			Spec. (%)			AUC			$\mathit{J}$ (Youden’s Index)
	LF	RF	CF	LF	RF	CF	LF	RF	CF	LF	RF	CF	LF	RF	CF
ALS vs. HC	100	100	100	100	100	100	100	100	100	1	1	1	1	1	1
HD vs. HC	98.41	98.04	97.56	98.54	97.59	98.51	98.25	98.60	96.41	0.9839	0.9810	0.9746	0.9679	0.9619	0.9492
PD vs. HC	100	100	100	100	100	100	100	100	100	1	1	1	1	1	1
ALS vs. HD	100	100	100	100	100	100	100	100	100	1	1	1	1	1	1
PD vs. ALS	95.64	95.95	94.21	94.07	94.59	92.95	97.63	97.65	95.78	0.9585	0.9612	0.9437	0.9170	0.9224	0.8873
HD vs. PD	97.11	97.25	94.98	96.81	96.54	93.54	97.51	98.24	97.14	0.9711	0.9739	0.9534	0.9432	0.9478	0.9068
NDD vs. HC	98.86	98.91	98.93	99.01	99.04	99.44	98.38	98.53	97.43	0.9870	0.9878	0.9844	0.9739	0.9757	0.9687

Note: bold and underlined were selected by Youden’s index criteria as the best classification result and model.