The Partial Least Squares (PLS) module is a comprehensive implementation of partial least squares regression analysis. Effects for a large number of predictors of any type on a large number of dependent (response) variables can be analyzed. Designs can include single-degree-of-freedom effects for continuous predictor variables, multiple-degrees-of-freedom effects for categorical predictor variables, or any combination of effects for continuous and categorical predictor variables. PLS implements partial least squares regression using the NIPALS (Rannar, Lindgren, Geladi, and Wold, 1994) and the SIMPLS (de Jong, 1993) algorithms for extracting partial least squares regression components.

The Introductory Overview topics listed below describe the use of partial least squares regression analysis. If you are unfamiliar with the basic methods of regression in linear models, it may be useful to first review the basic information on these topics in Elementary Statistical Concepts; the different designs available in this module are also described in the context of STATISTICA General Linear Model (GLM), Generalized Linear/Nonlinear Model (GLZ), and General Regression Models (GRM). To review the dialogs and results options, see the Partial Least Squares (PLS) Index.

Introductory Overview Topics

Training and Verification Samples