TABLE 9.

Graph features: performance and hyperparameters of graph models.

Model	Train_Acc	Val_Acc	Test_Acc	Train_F1	Val_F1	Test_F1	Hyperparameters
GraphsAGE_mean	92.67 ±0.2	87.71 ±0.111	84.61 ±0.621	0.955±0.004	0.943 ±0.006	0.881 ±0.011	hidden_dimensions=78, lr=0.01 ,dropout=0.309, Aggr=mean, num_layers=2
GraphSAGE_max	90.46 ±0.08	86.61 ±0.4243	83.42 ±0.825	0.909 ±0.005	0.880 ±0.003	0.868 ±0.005	hidden_dimensions=78, lr=0.01,dropout=0.309, Aggr=max, num_layers=2
GATV2	91.42 ±0.888	87.79 ±0.601	85.83±0.096	0.927 ±0.029	0.915 ±0.030	0.891±0.01	lr=le-2,heads=8, dropout=0.6, weight_decay=5e-4, num_layers=2
GatConv	88.35 ±1.266	85.08 ±1.53	84.12±1.03	0.874±0.041	0.852±0.03	0.8825 ±0.01	lr=0.0047, dropout=0.6928, heads=8, weight_decay=5e-4, num_layers=2
Proposed Model	92.88 ±0.021	88.47 ±0.62	87.21 ±0.98	0.9681 ±0.004	0.9603 ±0.005	0.9057 ±0.01	lr=0.01, hidden_channel=33, num_layers=4, dropout=0.2, heads=1, num_layers=4