Table 2.

Classification performance between groups (NC vs. iRBD and NC vs. PD) with two classification algorithm, logistic regression, and random forest. ACC = accuracy; SEN = sensibility; SPE = specificity, MDMR = multivariate distance matrix regression; NBS = network based statistics.

Connectivity features		MDMR			NBS			Whole Brain
		NC vs. PD	NC vs. iRBD	iRBD vs. PD	NC vs. PD	NC vs. iRBD	iRBD vs. PD	NC vs. PD	NC vs. iRBD	iRBD vs. PD
Logistic Regression	ACC	0.708	0.667	0.667	0.683	0.724	0.635	0.658	0.625	0.588
	SEN	0.750	0.667	0.701	0.634	0.725	0.548	0.607	0.705	0.550
	SPE	0.667	0.527	0.519	0.772	0.650	0.646	0.738	0.551	0.625
Random Forest	ACC	0.933	0.833	0.889	0.875	0.833	0.824	0.833	0.778	0.824
	SEN	0.967	0.886	0.925	0.917	0.875	0.869	0.857	0.765	0.812
	SPE	0.917	0.815	0.920	0.833	0.750	0.900	0.800	0.667	0.900