PLS - Partial Least Squares Regression (PLS) a very short...

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
Partial Least Squares Regression (PLS) – a very short introduction. PLS is a useful tool to clarify relationships between blocks of dependent variables (such as sensory descriptive data) and blocks of independent variables (such as volatile chemical analysis; vineyard data, etc.). It is a soft modeling technique, which permits the development of a model that describes the relationship between the two blocks or sets of data. It is not limited by the numbers of variables in either block relative to the number of cases, unlike PCA. In PCA you should always have more cases (wines) than you have variables (sensory terms). Essentially in a PLS analysis “factors” that are linear combinations of one data set (such as the chemical data) that predicts as much as possible of the variation in another data set (such as the descriptive ratings). By looking at the loadings of terms on the extracted factors, we can try to interpret the factors or dimensions. We can also see which terms in one block of data are “associated with”
Background image of page 1
This is the end of the preview. Sign up to access the rest of the document.

This note was uploaded on 09/29/2010 for the course VEN 91863 taught by Professor Hildergardheymann during the Spring '09 term at UC Davis.

Ask a homework question - tutors are online