Table 2.

Summary of average metrics scores for each aggregation method and classifier model, class size = 5.

Position	Dictionary	Model-Method	ACC	Precision	Recall	F1	BACC
1	Gabor-1000-20	LGBM-Mean	0.736	0.692	0.613	0.634	0.613
12	Gabor-125-20	XGBoost-Mean	0.729	0.676	0.613	0.632	0.613
39	KSVD-125-10	SVC-Mean	0.724	0.673	0.611	0.632	0.611
60	KSVD-250-5	LGBM-Max	0.72	0.668	0.624	0.641	0.624
93	KSVD-500-5	XGBoost-Max	0.715	0.664	0.614	0.633	0.614
115	KSVD-500-10	SVC-Max	0.709	0.667	0.573	0.593	0.573
150	Gabor-125-20	Random Forest-Mean	0.703	0.658	0.555	0.574	0.554
159	KSVD-250-20	MLP-Mean	0.703	0.635	0.613	0.622	0.613
187	Gabor-125-20	LGBM-Single	0.699	0.645	0.588	0.607	0.588
222	Gabor-125-20	XGBoost-Single	0.696	0.638	0.578	0.597	0.578
265	DL-1000-10	MLP-Max	0.691	0.616	0.563	0.578	0.563
373	KSVD-125-5	SVC-Single	0.681	0.631	0.551	0.573	0.551
393	Gabor-125-10	Random Forest-Single	0.678	0.636	0.539	0.562	0.539
459	KSVD-250-5	Random Forest-Max	0.671	0.651	0.522	0.549	0.522
564	KSVD-62-5	MLP-Single	0.656	0.566	0.556	0.560	0.556
696	Gabor-500-40	KNeighbors-Mean	0.632	0.597	0.468	0.482	0.468
698	DL-1000-20	GaussianNB-Mean	0.631	0.519	0.495	0.502	0.494
718	Gabor-125-5	AdaBoost-Mean	0.628	0.534	0.527	0.521	0.526
825	KSVD-250-5	Kneighbors-Max	0.61	0.629	0.43	0.440	0.430
865	DL-1000-10	GaussianNB-Max	0.603	0.498	0.465	0.473	0.465
943	Gabor-125-10	AdaBoost-Max	0.586	0.489	0.499	0.478	0.499
950	Gabor-125-10	AdaBoost-Single	0.585	0.485	0.486	0.472	0.486
971	Gabor-125-5	Decision Tree-Mean	0.582	0.474	0.476	0.474	0.476
977	Gabor-125-10	Kneighbors-Single	0.581	0.551	0.414	0.428	0.414
1086	Gabor-125-20	Decision Tree-Single	0.56	0.458	0.458	0.457	0.458
1150	DL-250-20	GaussianNB-Single	0.548	0.449	0.417	0.411	0.416
1176	KSVD-1000-5	Decision Tree-Max	0.543	0.447	0.44	0.443	0.439