Abstract

A near infrared calibration model with higher precision and better stability was constructed in the present study, using 280 cottonseed samples. The reference phytic acid contents were determined by high-performance ion chromatography. A combination of Savitzky–Golay smoothing, standard normal variate, and the first derivative was chosen as the spectral pre-treatment method. Monte Carlo uninformative variable elimination was proposed for spectral variable selection. The regression methods of partial least squares, least squares support vector machines, and weighted least squares support vector machines were developed for the calibration model. The optimal near infrared calibration model for phytic acid contents in the cottonseed meals was least squares support vector machines, with r2 = 0.97, RPD = 5.53, RMSECV = 0.06%, and RMSEP = 0.05%. This robust method can replace the traditional method of phytic acid determination in cottonseed meals.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call