Results of applying first-order algorithms to validation samples.

Model	LVsa	PPb	δ	PDCc	RMSECV	RMSEP	REP (%)	R 2	Q 2
PLS-1	3	—	—	—	0.0121	0.0144	15.3069	0.6921	0.7001
MC-PLS-1	2	—	—	—	0.0044	0.0067	7.1529	0.8120	0.8097
ASC-PLS-1	2	—	—	—	0.0033	0.0064	6.8227	0.8199	0.8123
CPR	2	1	—	—	0.0011	0.0062	6.5313	0.8341	0.8411
MC-CPR	1	1	—	—	0.0021	0.0034	3.6328	0.8540	0.8548
ASC-CPR	2	0.5	—	—	0.0088	0.0064	6.8227	0.8500	0.8533
RMC-PRM	1	—	—	10	0.0010	0.0002	0.6004	0.9121	0.9230
RLMC-PRM	1	—	—	10	0.000021	0.00001	0.0102	0.9891	0.9901
RMC-RCR	1	—	0.5	10	0.0038	0.002	2.1321	0.9231	0.9199
RLMC-RCR	1	—	0.5	10	0.0042	0.002	2.1317	0.9188	0.9321
MLR-SPA	1	—	—	—	0.2109	0.1931	204.97	0.5411	0.4830

^aLatent variables.

^bPower parameter.

^cPercentage of data contamination.