A method of two-dimensional correlation spectroscopy combined with residual neural network for comparison and differentiation of medicinal plants raw materials superior to traditional machine learning: a case study on Eucommia ulmoides leaves

Table 2 The classification parameters of PLS-DA established for different regions and drying methods

Styles	Classes	Train set				Test set
Styles	Classes	SEN	SPE	ACC	ROC	SEN	SPE	ACC	ROC
different regions	Guizhou	0.8117	0.7124	0.8978	0.9444	0.9722	0.8880	0.9012	0.9335
	Henan	0.9405	0.7446	0.9570	0.9805	0.8857	0.9685	0.9506	0.9852
	Hubei	0.7556	0.8844	0.9489	0.9018	0.7368	0.9720	0.9444	0.9653
	Hunan	0.7885	0.8414	0.9516	0.9202	0.6296	0.1000	0.9383	0.9358
	Jiangxi	0.8519	0.9167	0.9785	0.9794	0.5000	0.1000	0.9630	09,978
	Shaanxi	0.7458	0.8118	0.9301	0.9907	0.1000	0.9648	0.1000	0.9527
	Xinjiang	0.9500	0.8925	0.9167	0.1000	0.8750	0.1000	0.9938	0.1000
Different processing	40°C	0.8000	0.9315	0.8279	0.8991	0.8537	0.9160	0.9000	0.9418
	60°C	0.9677	0.9609	0.9677	0.9934	0.9750	0.9664	0.9625	0.9967
	Sun drying	0.2796	0.9395	0.7754	0.7175	0.6667	0.9308	0.8563	0.7404
	Shade drying	0.6667	0.8398	0.8011	0.8257	0.7949	0.8678	0.8500	0.8873

SEN Sensitivity, SPE specificity, ACC Accuracy, ROC Receiver Operating Characteristic

ISSN: 1746-4811