Partial least squares for discrimination
Tóm tắt
Partial least squares (PLS) was not originally designed as a tool for statistical discrimination. In spite of this, applied scientists routinely use PLS for classification and there is substantial empirical evidence to suggest that it performs well in that role. The interesting question is: why can a procedure that is principally designed for overdetermined regression problems locate and emphasize group structure? Using PLS in this manner has heurestic support owing to the relationship between PLS and canonical correlation analysis (CCA) and the relationship, in turn, between CCA and linear discriminant analysis (LDA). This paper replaces the heuristics with a formal statistical explanation. As a consequence, it will become clear that PLS is to be preferred over PCA when discrimination is the goal and dimension reduction is needed. Copyright © 2003 John Wiley & Sons, Ltd.
Từ khóa
Tài liệu tham khảo
Gottfries J, 1995, Diagnosis of dementias using partial least squares discriminant analysis, Dementia, 6, 83
Delwiche S, 1972, Differentiation of hard red wheat by near‐infrared analysis of bulk samples, Grain Qual., 3, 243
Helland IS, 1992, Maximum likelihood regression on relevant components, J. R. Statist. Soc. B, 54, 637
Wold H, 1966, Multivariate Analysis, 391
Wold H, 1981, Systems under Indirect Observation, Causality–Structure–Prediction, 1
BarkerM.Partial least squares for discrimination.PhD Dissertation University of Kentucky 2000.
HinkleJandRayensWS.Partial least squares reciprocal components and reciprocal curves.Proceedings from the 1998 Joint Statistical Meetings Dallas TX 1998;126–130.
Stone M, 1990, Continuum regression: cross‐validated sequentially constructed prediction embracing ordinary least squares, partial least squares, and principal component regression, J. R. Statist. Soc. B, 52, 237
HinkleJ.Reciprocal components reciprocal curves and partial least squares.PhD Dissertation University of Kentucky 1995.
Rayens WS, 2003, Oriented partial least squares, Ital. J. Appl. Statist.
RayensWS.The art of maximizing covariance.University of Kentucky Technical Report 383 2000.