Table 3. Correctly classified percentages of grapevine leaves for each Savitzky-Golay filter and algorithm combination for N = 20 and N = 5 number of varieties.

Number of varieties	Savitzky-Golay filter	PLS-DA	SVM	ANN	significance
N = 20	D1W5	45.3 B	77.5 A a	81.2 A b	* * *
	D1W11	44.4 B	69.7 A b	67.3 A c	* * *
	D2W5	51.0 B	78.0 A a	84.3 A a	* * *
	D2W11	43.2 B	76.7 A a	79.8 A b	* * *
	significance	n.s.	* *	* * *
N = 5	D1W5	81.5 a	87.8 b	88.0 b	n.s.
	D1W11	72.5 B b	79.3 A c	81.0 A c	*
	D2W5	84.5 B a	91.2 A a	91.6 A a	* * *
	D2W11	66.0 B c	87.8 A b	86.9 A b	* * *
	significance	* *	* * *	* * *

The values shown are the varieties correctly classified percentage. Each value is, in turn, the average of the results obtained using and not using scatter correction and, for SVM and ANN, the 12 parameter sets.

PLS-DA: Partial Least Squares Discriminant Analysis; SVM: Support Vector Machine; ANN: Artificial Neural Network.

Uppercase and italic lowercase letters attend respectively to row-wise (comparison among algorithms) and column-wise (comparison among Savitzky-Golay filters) values comparison. n.s.: not significant (p ≥ 0.05); *: p < 0.05; * *: p < 0.01; * * *: p < 0.001.