PoLiSh - smoothed partial least-squares regression

Citation
Dn. Rutledge et al., PoLiSh - smoothed partial least-squares regression, ANALYT CHIM, 446(1-2), 2001, pp. 281-296
Citations number
12
Categorie Soggetti
Spectroscopy /Instrumentation/Analytical Sciences
Journal title
ANALYTICA CHIMICA ACTA
ISSN journal
00032670 → ACNP
Volume
446
Issue
1-2
Year of publication
2001
Pages
281 - 296
Database
ISI
SICI code
0003-2670(20011119)446:1-2<281:P-SPLR>2.0.ZU;2-Y
Abstract
Partial least-squares (PLS) regression is a very widely used technique in s pectroscopy for calibration/prediction purposes. One of the most important steps in the application of the PLS regression is the determination of the correct number of dimensions to use in order to avoid over-fitting, and the refore to obtain a robust predictive model. The "structured" nature of spec troscopic signals may be used in several ways as a guide to improve the PLS models. The aim of this work is to propose a new technique for the applica tion of PLS regression to signals (FT-IR, NMR, etc.). This technique is bas ed on the Savitsky-Golay (SG) smoothing of the loadings weights vectors (w) obtained at each iteration step of the NIPALS procedure. This smoothing pr ogressively "displaces" the random or quasi-random variations from earlier (most important) to later (less important) PLS latent variables. The Durbin -Watson (DW) criterion is calculated for each PLS vectors (p, w, b) at each iteration step of the smoothed NIPALS procedure in order to measure the ev olution of their "noise" content. PoLiSh has been applied to simulated data sets with different noise levels and it was found that for those with noise levels higher than 10-20%, an improvement in the predictive ability of the models is observed. This technique is also important as a tool to evaluate the true dimensionality of signal matrices for complex PLS models, by comp aring the DW profile of the PoLiSh vectors at different smoothing degrees w ith those of the unsmoothed PLS models. (C) 2001 Elsevier Science B.V. All rights reserved.