Premium
Comparing and predicting sensory profiles from NIRS data: use of the GOMCIA and GOMCIA‐PLS multiblock methods
Author(s) -
Vivien Myrtille,
Verron Thomas,
Sabatier Robert
Publication year - 2005
Publication title -
journal of chemometrics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.47
H-Index - 92
eISSN - 1099-128X
pISSN - 0886-9383
DOI - 10.1002/cem.919
Subject(s) - partial least squares regression , calibration , repeatability , regression , sensory system , pattern recognition (psychology) , computer science , regression analysis , linear regression , mathematics , artificial intelligence , statistics , psychology , cognitive psychology
Sequential multiblock component and regression methods are used to analyse pea sensory data. First the repeatability of sensory profiles is studied with the generalized orthogonal multiple co‐inertia analyis (GOMCIA) method. Then a calibration model is built to predict the averaged sensory profiles from near‐infrared spectroscopy (NIRS) data by means of the GOMCIA partial least squares (GOMCIA‐PLS) method. It is proposed to split the NIRS data into blocks and search the most influential block in the calibration model in order to select spectral wavelengths. This approach is compared with a PLS regression model. It is shown that a PLS regression model can be less efficient than a multiblock regression model in terms of prediction. Copyright © 2005 John Wiley & Sons, Ltd.