Analysis of Artificial Cow-Bezoar by Near-Infrared Spectroscopy Coupled with Competitive Adaptive Reweighted Sampling Method / 中国药学杂志

ABSTRACT

OBJECTIVE: To analyze and study the characteristic variables of wavelength in near-infrared spectroscopy of artificial cow-bezoar. METHODS: A method of near-infrared spectroscopy coupled with competitive adaptive reweighted sampling(CARS) was performed in characteristic variables of wavelength screening for the qualitative and quantitative researches, respectively. RESULTS: Some characteristic variables of wavelength, 0.48%-4.44% of all variables of wavelength, were screened out by CARS for different models. Not only the number of variables for building models decreased significantly, but also the index parameters for evaluating model became better. CONCLUSION: This method is suitable for quality evaluation and quality control for artificial cow-bezoar.