Click on “Download PDF” for the PDF version or on the title for the HTML version.


If you are not an ASABE member or if your employer has not arranged for access to the full-text, Click here for options.

Brand classification and storage time prediction of commercial infant milk formula powder using NIR spectroscopy and chemometrics

Published by the American Society of Agricultural and Biological Engineers, St. Joseph, Michigan www.asabe.org

Citation:  2021 ASABE Annual International Virtual Meeting  2101066.(doi:10.13031/aim.202101066)
Authors:   Renxi Kang, Xiao Wang, Lisa E Henihan, Colm P O’Donnell
Keywords:   brand classification, chemometrics, commercial infant milk formula (CIMF) powder, near infrared spectroscopy (NIRS), storage time prediction.

Abstract. This study investigated the feasibility of using near infrared spectroscopy (NIRS) over a spectral range of 1158-2169 nm with chemometrics for (i) classification of commercial infant milk formula (CIMF) powder by brand (i.e., Brand A, B and C), and (ii) prediction of CIMF powder storage time (i.e., 0 -12 months). For classification of CIMF powder by brand, data driven soft independent method of class analogy (DD-SIMCA) models developed had sensitivities of 88.0 % - 98.0 % and specificities of 53 % - 100 % for prediction. Partial least square discrimination analysis (PLSDA) models developed had sensitivities, specificities and efficiencies of 32 % - 100 %, 90 % - 100 % and 56 % - 100 % for prediction. Hard PLSDA had overall better performances than soft PLSDA models. Partial least square regression (PLSR) modelling and variable importance on projection (VIP) were employed to develop CIMF powder storage time prediction models, and had coefficients of determination on prediction (R2Ps) of 0.93 - 0.97, root mean square errors of prediction (RMSEPs) of 0.76 - 1.15 months and ratio of prediction error to deviation in prediction (RPDPs) of 3.69 - 7.18. This study demonstrated the feasibility of using NIRS combined with chemometrics for CIMF powder brand classification and storage time prediction during CIMF powder manufacture and along the supply chain.

(Download PDF)    (Export to EndNotes)