Table 2.

Performance (mean ± SD) of different models in autism spectrum disorder (ASD) vs. normal control (NC) classification based on resting-state functional MRI (rs-fMRI) data in NYU site of the ABIDE dataset.

Model	Accuracy	Recall	Precision	F1-score	AUC
DCF	0.624 ± 0.060	0.703 ± 0.197	0.676 ± 0.142	0.652 ± 0.076	0.711 ± 0.071
LCCF	0.705 ± 0.085	0.676 ± 0.196	0.819 ± 0.171	0.698 ± 0.102	0.812 ± 0.033
CCF	0.672 ± 0.111	0.781 ± 0.153	0.724 ± 0.173	0.719 ± 0.060	0.792 ± 0.039
GCNA	0.758 ± 0.075	0.761 ± 0.146	0.789 ± 0.089	0.763 ± 0.087	0.854 ± 0.058
GCNC	0.753 ± 0.080	0.773 ± 0.171	0.759 ± 0.100	0.757 ± 0.120	0.808 ± 0.074
MGRL (Ours)	0.795 ± 0.068	0.809 ± 0.146	0.823 ± 0.103	0.802 ± 0.075	0.886 ± 0.062

The bold values mean to highlight the experiment results.