Table 5.

Comparison of SDL vs. HDL vs. EDL models.

Comparison for six binary classifiers and four multiclass classifiers of SDL, HDL and EDL Models
Dataset	SDL		HDL		EDL		Absolute difference (%)
	${\mathrm{\eta }}_{\text{SDL}}$(%)	${\mathrm{\alpha }}_{\text{SDL}}$[0–1]	${\mathrm{\eta }}_{\text{HDL}}$(%)	${\mathrm{\alpha }}_{\text{HDL}}$[0–1]	${\mathrm{\eta }}_{\text{EDL}}$(%)	${\mathrm{\alpha }}_{\text{EDL}}$[0–1]	${\mathrm{\eta }}_{\text{D1}}$	${\mathrm{\alpha }}_{\text{D1}}$	${\mathrm{\eta }}_{\text{D2}}$	${\mathrm{\alpha }}_{\text{D2}}$	${\mathrm{\eta }}_{\text{D3}}$	${\mathrm{\alpha }}_{\text{D3}}$
Binary Class (BC) Classification
Human vs. Gorilla	89.31	0.8998	92.19	0.9336	95.29	0.9575	2.88	3.38	3.1	2.39	5.98	5.77
Human vs. Mouse	81.21	0.8203	82.99	0.8384	85.77	0.8636	1.78	1.81	2.78	2.52	4.56	4.33
Human vs. Rat	83.19	0.8397	85.57	0.8766	87.92	0.8851	2.38	3.69	2.35	0.85	4.73	4.54
Mouse vs. Gorilla	94.11	0.9518	96.68	0.9704	97.28	0.9737	2.57	1.86	0.6	0.33	3.17	2.19
Mouse vs. Rat	91.23	0.9212	94.3	0.9575	95.66	0.9735	3.07	3.63	1.36	1.6	4.43	5.23
Rat vs. Gorilla	92.31	0.9295	95.38	0.9611	96.03	0.9696	3.07	3.16	0.65	0.85	3.72	4.01
Mean of 6 BC	88.56	0.8937	91.19	0.923	92.99	0.9372	2.63	2.93	1.8	1.42	4.43	4.35
Multiclass (MC) Classification
Human vs. All	85.1	0.8545	85.88	0.8738	88.65	0.8934	0.78	1.93	2.77	1.96	3.55	3.89
Gorilla vs. All	93.53	0.9421	95.95	0.9702	97.45	0.9765	2.42	2.81	1.5	0.63	3.92	3.44
Rat vs. All	89.43	0.9104	90.54	0.9174	93.06	0.939	1.11	0.7	2.52	2.16	3.63	2.86
Mouse vs. All	87.68	0.8947	89.3	0.9046	91.74	0.924	1.62	0.99	2.44	1.94	4.06	2.93
Mean of 4 MCC	88.94	0.9004	90.42	0.9165	92.73	0.9332	1.48	1.61	2.31	1.67	3.79	3.28
Binary class + Multiclass Classification
Mean of 10 Classifiers	88.71	0.8964	90.88	0.9204	92.89	0.9356	2.17	2.4	2.01	1.52	4.18	3.92

$\mathrm{\eta }$ (%) represents accuracy and $\mathrm{\alpha }$ (0-1) represents AUC. ${\mathrm{\eta }}_{\text{SDL}}$: Mean accuracy of SDL models; ${\mathrm{\alpha }}_{\text{SDL}}$: Mean AUC of SDL models; ${\mathrm{\eta }}_{\text{HDL}}$: Mean accuracy of HDL models; ${\mathrm{\alpha }}_{\text{HDL}}$: Mean AUC of HDL models; ${\mathrm{\eta }}_{\text{EDL}}$: Mean accuracy of EDL models; ${\mathrm{\alpha }}_{\text{EDL}}$: Mean AUC of EDL models;${\mathrm{\eta }}_{\text{D1}}$: Mean absolute accuracy difference (HDL vs. SDL); ${\mathrm{\eta }}_{\text{D1}}\left(\text{\%}\right)=\left|{\mathrm{\eta }}_{\text{HDL}}-{\mathrm{\eta }}_{\text{SDL}}\right|$; ${\mathrm{\alpha }}_{\text{D1}}$ : Mean absolute AUC difference (HDL vs. SDL); ${\mathrm{\alpha }}_{\text{D1}}\text{(\% ) =}\left|{\mathrm{\alpha }}_{\text{HDL}}-{\mathrm{\alpha }}_{\text{SDL}}\right|\times 100$; ${\mathrm{\eta }}_{\text{D2}}$: Mean absolute accuracy difference (EDL vs. HDL); ${\mathrm{\eta }}_{\text{D2}}\text{(\% )}=\left|{\mathrm{\eta }}_{\text{EDL}}-{\mathrm{\eta }}_{\text{HDL}}\right|$ ${\mathrm{\alpha }}_{\text{D2}}$: Mean absolute AUC difference (EDL vs. HDL); ${\mathrm{\alpha }}_{\text{D2}}\text{(\% ) =}\left|{\mathrm{\alpha }}_{\text{EDL}}-{\mathrm{\alpha }}_{\text{HDL}}\right|\times 100$; ${\mathrm{\eta }}_{\text{D3}}$: Mean absolute accuracy difference (EDL vs. SDL); ${\mathrm{\eta }}_{\text{D3}}\text{(\% )}=\left|{\mathrm{\eta }}_{\text{EDL}}-{\mathrm{\eta }}_{\text{SDL}}\right|$; ${\mathrm{\alpha }}_{\text{D3}}$: Mean absolute AUC difference (EDL vs. SDL); ${\mathrm{\alpha }}_{\text{D3}}\text{(\% ) =}\left|{\mathrm{\alpha }}_{\text{EDL}}-{\mathrm{\alpha }}_{\text{SDL}}\right|\times 100$.

Significant values are in [bold].