Determination of active ingredients in matrine aqueous solutions by mid-infrared spectroscopy and competitive adaptive reweighted sampling

Abstract

In this paper, we use MIR and variable selection techniques to determine the active ingredients of matrine aqueous solutions. Competitive adaptive reweighted sampling (CARS) was engaged to investigate effective variables and compared with full-spectrum partial least squares (PLS) regression. Results indicate that MIR is very efficient in aqueous solutions quantitative determination. In the meantime, CARS is proved to be an effective variable selection approach to eliminate spectral redundancy.

Publication

Optik

Type

Journal Articles