The L-Curve Criterion as a Model Selection Tool in PLS Regression

Table 5

The result of comparison between cross validation (CV), generalized cross validation (GCV), and L-curve criterion in PLS regression in terms of the number of components retained in the final model.

	The number of PLS components retained in the final model
	CV	GCV	L-curve criterion
(m, 0.5, 1)	1	6	4
(m, 0.5, 5)	1	6	4
(m, 0.5, 7)	2	6	4
(m, 0.8, 1)	1	6	3
(m, 0.8, 5)	1	6	4
(m, 0.8, 7)	1	10	5

A negative value of .