PARTIAL LEAST-SQUARES MODELING AS SUCCESSIVE SINGULAR-VALUE DECOMPOSITIONS

The vectors p, q and w of the partial least squares (PLS) model are shown to be computable from the input covariance matrix and the input/output covariance matrix of a system without reference to any actual data. This approach is given in terms of the singular value decompositions of the input/output covariance matrix and its residual matrices for each component. Such an approach may be useful for studying the properties of the PLS model.