“…GPA generates a consensus plot that matches the configuration of the data sets by centering, rotating and adjusting scales of each data set (Chung et al, 2003). Thus, the three major elements, samples, panelists and sensory attributes, can be examined and contrasted using a single encompassing analysis (de Jong, Heidema, & van der Knaap, 1997). Partial least squares regression (PLSR) is frequently used to understand the relationships between instrumental (X) and sensory (Y) data sets by predicting one data set from the other .…”