A Dataset on Corn Silage in China Used to Establish a Prediction Model Showing Variation in Nutrient Composition

Table 7

Comparison of performance of PLSR and the BP-ANN algorithm for predicting nutrient contents of whole plant corn silage.

Nutrition index	Model	Calibration set		Prediction set
		R^2c	RMSEC	R^2p	RMSEP
DM	PLSR	0.836	0.675	0.823	0.701
	BP-ANN	0.900	0.550	0.845	0.642
CP	PLSR	0.903	0.407	0.900	0.418
	BP-ANN	0.945	0.293	0.927	0.519
EE	PLSR	0.788	0.261	0.763	0.276
	BP-ANN	0.886	0.420	0.853	0.579
Ash	PLSR	0.795	0.543	0.779	0.564
	BP-ANN	0.902	0.806	0.847	0.792
NDF	PLSR	0.921	1.756	0.916	1.813
	BP-ANN	0.965	0.650	0.935	0.627
ADF	PLSR	0.910	1.281	0.904	1.321
	BP-ANN	0.991	0.667	0.975	0.993
Starch	PLSR	0.933	1.952	0.929	2.004
	BP-ANN	0.972	0.763	0.944	0.761
Ca	PLSR	0.418	0.061	0.376	0.064
	BP-ANN	0.730	0.806	0.509	0.792
	PLSR	0.499	0.029	0.477	0.030
	BP-ANN	0.615	0.806	0.453	0.792