Table 3.

Predictive MAE, MASE, RMSE, and R² for the DL architecture for BW, MD, and BF for the three genetic lines evaluated on different validation schemes

Trait	Validation Schemes1	Line	MAE	MASE	RMSE	R 2
BW, kg	Line	1	4.86	4.22	6.34	0.68
	Line	2	4.27	3.57	6.05	0.67
	Line	3	4.33	3.38	5.40	0.71
	3F-CV	1	4.26	3.70	5.43	0.78
	3F-CV	2	3.99	3.35	5.70	0.68
	3F-CV	3	3.58	2.80	4.44	0.81
	5F-CV	1	3.46	3.02	4.74	0.81
	5F-CV	2	3.29	2.77	4.95	0.78
	5F-CV	3	2.78	2.17	3.58	0.87
MD, mm	Line	1	4.17	6.84	5.22	0.13
	Line	2	5.20	7.63	6.43	0.05
	Line	3	4.04	6.17	4.95	0.10
	3F-CV	1	4.45	7.08	5.40	0.10
	3F-CV	2	5.15	7.75	6.44	0.04
	3F-CV	3	4.04	6.17	5.00	0.08
	5F-CV	1	3.39	5.50	4.38	0.36
	5F-CV	2	3.57	5.27	4.83	0.39
	5F-CV	3	3.04	4.62	4.00	0.41
BF, mm	Line	1	0.97	15.15	1.22	0.20
	Line	2	1.22	20.47	1.55	0.03
	Line	3	1.20	19.80	1.47	0.06
	3F-CV	1	1.00	16.43	1.22	0.22
	3F-CV	2	1.12	18.69	1.41	0.19
	3F-CV	3	1.11	18.03	1.39	0.17
	5F-CV	1	0.81	13.16	1.05	0.41
	5F-CV	2	0.85	14.19	1.15	0.47
	5F-CV	3	0.89	14.39	1.21	0.36

¹Retaining one line for validation (Line), a 3-fold (3F-CV), and a 5-fold (5F-CV) CV.