Premium
Panelists bias matrix estimation in a red wine trained panel: A potential tool for data pre‐treatment and feedback calibration
Author(s) -
Diako Charles,
Cooper Kevin D.,
Ross Carolyn F.
Publication year - 2019
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.3084
Subject(s) - wine , statistics , matrix (chemical analysis) , filter (signal processing) , mathematics , calibration , computer science , econometrics , artificial intelligence , computer vision , optics , materials science , physics , composite material
In this study, panelists' bias in a trained wine sensory panel was abstracted as a linear operator to provide feedback to the panelists, to adjust (“filter”) evaluations and to predict attribute ratings of unknown samples. The bias matrix is a square matrix of attributes showing individual panelists intensity agreement with the entire panel average. Using this matrix to filter the original intensity ratings reduced the dispersion of ratings around the mean without significantly affecting the mean value. Predictively filtered means were closer to bias‐adjusted means than the un‐adjusted means.