. 2019 Oct 18;8(10):511. doi: 10.3390/foods8100511

Table 1.

Evaluation of models obtained by orthogonal partial least squares (OPLS).

Sample		Scaling	A ^a	N ^b	R²X	R²Y	Q²Y	Regression Line
Sample		Scaling	A ^a	N ^b	R²X	R²Y	Q²Y	y	R²	RMSEE	RMSEcv
Dark muscle	0 °C	UV	1+0+0	12	0.442	0.742	0.668	0.983x + 0.683	0.755	2.920	3.101
	0 °C	Par	1+1+0	12	0.806	0.727	0.535	0.950x + 0.551	0.732	3.164	3.683
	5 °C	UV	1+1+0	12	0.616	0.952	0.865	1.008x + 0.114	0.954	0.680	1.111
	5 °C	Par	1+4+0	12	0.941	0.959	0.909	0.996x − 0.036	0.960	0.765	0.949
Ordinary muscle	0 °C	UV	1+1+0	12	0.521	0.874	0.611	0.997x + 0.707	0.891	2.151	3.471
	0 °C	Par	1+2+0	12	0.722	0.898	0.701	0.999x + 0.016	0.898	2.052	2.865
	5 °C	UV	1+1+0	12	0.578	0.884	0.778	0.973x + 0.375	0.898	1.054	1.455
	5 °C	Par	1+5+0	12	0.981	0.995	0.955	1.002x − 0.050	0.995	0.294	0.659

^aA = number of models. ^b N = number of samples used in producing models. UV = unit variance-scaling; Par = Pareto-scaling; RMSEE = root mean square errors of estimation; and RMSEcv: root mean square errors of cross-validation.