Table 3.

Performance (mean ± SD) of different models in ASD vs. NC classification based on rs-fMRI data in UM site of the ABIDE dataset.

Model	Accuracy	Recall	Precision	F1-score	AUC
DCF	0.645 ± 0.089	0.733 ± 0.187	0.691 ± 0.046	0.702 ± 0.103	0.633 ± 0.039
LCCF	0.620 ± 0.077	0.720 ± 0.200	0.670 ± 0.059	0.681 ± 0.116	0.573 ± 0.045
CCF	0.634 ± 0.105	0.761 ± 0.190	0.662 ± 0.076	0.702 ± 0.131	0.542 ± 0.049
GCNA	0.739 ± 0.098	0.710 ± 0.159	0.927 ± 0.067	0.790 ± 0.102	0.875 ± 0.065
GCNC	0.725 ± 0.082	0.651 ± 0.186	0.751 ± 0.108	0.677 ± 0.121	0.824 ± 0.057
MGRL (Ours)	0.762 ± 0.078	0.843 ± 0.070	0.794 ± 0.100	0.812 ± 0.054	0.867 ± 0.049

The bold values mean to highlight the experiment results.