Table 28.

Comparison of the BPSO-kNN with kNN*, CNN, and MLP. The table compares the performance of the proposed model, BPSO-kNN, which incorporates feature selection, with other models, including kNN*, CNN, and MLP, which do not employ feature selection.

Dataset	CNN		MLP		kNN*		BPSO-kNN
	Accuracy	Time	Accuracy	Time	Accuracy	Time	Accuracy	Time
All	0.6105	291.656	0.5438	0.789	0.7409	1.706	0.7628	464.050
Caucasion	0.7283	204.591	0.6159	3.282	0.7483	0.177	0.8080	368.142
Hispanic	0.6513	180.510	0.5285	2.341	0.7724	0.194	0.7968	245.754
Females	0.7023	208.605	0.6102	3.398	0.7373	0.165	0.8136	266.882
Males	0.6263	174.349	0.5769	2.969	0.6987	0.225	0.7885	260.481
Age ≤ 50	0.6427	177.557	0.5780	2.917	0.7431	0.176	0.8624	261.377
Age > 50	0.6629	219.354	0.5455	3.483	0.7333	0.175	0.8061	382.676