Integration of Artificial Neural Network Modeling and Hyperspectral Data Preprocessing for Discrimination of Colla Corii Asini Adulteration

Table 1

Performance of PNN and GRNN with different preprocessing methods.

Modeling methods	Training set	Prediction set	CCR
PNN	120	80	82.5%
PNN + SG	120	80	85.0%
PNN + MSC	120	80	87.5%
PNN + MSC-SG	120	80	90.0%
GRNN	120	80	82.5%
GRNN + SG	120	80	87.5%
GRNN + MSC	120	80	87.5%
GRNN + MSC-SG	120	80	92.5%