Calibration of artificial neural network and partial least squares regression models for the prediction of secoisolariciresinol diglucoside contents in microwave-assisted extracts of various flaxseed (Linum usitatissimum L.) samples

Abstract

Abstract Partial least squares (PLS) and artificial neural network (ANN) regression models were calibrated for predicting the content of secoisolariciresinol diglucoside (SDG) in six flaxseed cultivars, defatted flaxseed meal and flax hulls extracts. The SDG was quantified by HPLC after microwave-assisted extraction (MAE) from flaxseed; the data were used in conjunction with the light absorption of the extracts measured after Folin–Ciocalteu’s assay at 289, 298, 343 and 765 nm, in order to calibrate the predictive PLS and ANN models. The accuracy and the predictive ability of the models ranged from good to excellent as indicated by RPD values (the ratio of the standard deviation of the reference values to the standard error of prediction) of 5.03–13.7. The PLS and ANN predictive models are useful to the flaxseed processing industry for rapidly and accurately predicting the SDG contents of various flaxseed samples based on their UV–Vis light absorption.