A Variable Selection Method Based on Fast Nondominated Sorting Genetic Algorithm for Qualitative Discrimination of Near Infrared Spectroscopy

Table 1

Number of calibration and prediction samples in each class.

Class	Calibration set	Prediction set
B	33	10
C	30	12
X	30	9
Total	93	31

Note. B denotes upper tobacco leaf, C denotes middle tobacco leaf, and X denotes lower tobacco leaf.