De nombreuses variables d’intérêt, comme par exemple des résultats expérimentaux, des rendements ou des indicateurs économiques, s’expriment naturellement sous la forme de taux, de proportions ou d’indices dont les valeurs sont nécessairement comprises entre zéro et un ou plus généralement deux valeurs fixes connues à l’avance. La régression Bêta permet de modéliser ces données avec beaucoup de souplesse puisque les fonctions de densité des lois Bêta peuvent prendre des formes tr ès variées. Toutefois, comme tous les mod èles de régression usuels, elle ne peut s’appliquer directement lorsque les prédicteurs présentent des probl èmes de multicolinéarité ou pire lorsqu’ils sont plus nombreux que les observations. Ces situations se rencontrent fréquemment de la chimie à la médecine en passant par l’économie ou le marketing. Pour circonvenir cette difficulté, nous formulons une extension de la régression PLS pour les mod èles de régression Bêta. Celle-ci, ainsi que plusieurs outils comme la validation croisée et des techniques bootstrap, est disponible pour le langage R dans la biblioth èque plsRbeta.
Many responses, for instance experimental results, yields or economic indices, can be naturally expressed as rates or proportions whose values must lie between zero and one or between any two given values. The Beta regression often allows to model these data accurately since the shapes of the densities of Beta laws are very versatile. Yet, as any of the usual regression model, it cannot be applied safely in case of multicollinearity and not at all when the model matrix is rectangular. These situations are frequently found from chemistry to medicine through economics or marketing. To circumvent this difficulty, we derived an extension of PLS regression to Beta regression models. It, as well as several other tools, such as cross validation or bootstrap techniques, is available for the R language in the plsRbeta package.
Keywords: Beta Regression, PLS Regression, PLS Beta Regression, Cross validation, Bootstrap techniques, R language
@article{JSFS_2013__154_3_143_0, author = {Bertrand, Fr\'ed\'eric and Meyer, Nicolas and Beau-Faller, Mich \`ele and El Bayed, Karim and Namer, Izzie-Jacques and Maumy-Bertrand, Myriam}, title = {R\'egression {B\^eta} {PLS}}, journal = {Journal de la soci\'et\'e fran\c{c}aise de statistique}, pages = {143--159}, publisher = {Soci\'et\'e fran\c{c}aise de statistique}, volume = {154}, number = {3}, year = {2013}, zbl = {1316.62179}, language = {fr}, url = {http://www.numdam.org/item/JSFS_2013__154_3_143_0/} }
TY - JOUR AU - Bertrand, Frédéric AU - Meyer, Nicolas AU - Beau-Faller, Mich èle AU - El Bayed, Karim AU - Namer, Izzie-Jacques AU - Maumy-Bertrand, Myriam TI - Régression Bêta PLS JO - Journal de la société française de statistique PY - 2013 SP - 143 EP - 159 VL - 154 IS - 3 PB - Société française de statistique UR - http://www.numdam.org/item/JSFS_2013__154_3_143_0/ LA - fr ID - JSFS_2013__154_3_143_0 ER -
%0 Journal Article %A Bertrand, Frédéric %A Meyer, Nicolas %A Beau-Faller, Mich èle %A El Bayed, Karim %A Namer, Izzie-Jacques %A Maumy-Bertrand, Myriam %T Régression Bêta PLS %J Journal de la société française de statistique %D 2013 %P 143-159 %V 154 %N 3 %I Société française de statistique %U http://www.numdam.org/item/JSFS_2013__154_3_143_0/ %G fr %F JSFS_2013__154_3_143_0
Bertrand, Frédéric; Meyer, Nicolas; Beau-Faller, Mich èle; El Bayed, Karim; Namer, Izzie-Jacques; Maumy-Bertrand, Myriam. Régression Bêta PLS. Journal de la société française de statistique, Méthodes statistiques en agronomie, Tome 154 (2013) no. 3, pp. 143-159. http://www.numdam.org/item/JSFS_2013__154_3_143_0/
[1] Two graphical displays for outlying and influential observations in regression, Biometrika, Volume 68 (1981) no. 1, pp. 13-20 | Zbl
[2] Plots, Transformations and Regression : An Introduction to Graphical Methods of Diagnostic Regression Analysis, Oxford University Press, Oxford, 1985 | Zbl
[3] Deviance residuals based PLS regression for censored data in high dimensional setting, Chemometrics and Intelligent Laboratory Systems, Volume 91 (2008) no. 1, pp. 78-86
[4] PLS generalised linear regression, Computational Statistics & Data Analysis, Volume 48 (2005) no. 1, pp. 17-46 | Zbl
[5] Beta Regression in R, Journal of Statistical Software, Volume 34 (2010) no. 2, pp. 1-24
[6] boot : Bootstrap R (S-Plus) Functions (2009) (R package version 1.2-37)
[7] Bootstrap confidence intervals (with Discussion), Statistical Science, Volume 11 (1996), pp. 189-228 | Zbl
[8] Bootstrap methods and their application, Cambridge University Press, Cambridge, 1997 | Zbl
[9] An Introduction to the Bootstrap, Chapman & Hall, New York, 1993 | Zbl
[10] Beta Regression for Modeling Rates and Proportions, Journal of Applied Statistics, Volume 31 (2004) no. 7, pp. 799-815 | Zbl
[11] Extended Beta Regression in R : Shaken, Stirred, Mixed and Partitioned, Journal of Statistical Software, Volume 48 (2012) no. 11, pp. 1-25
[12] PLS regression methods, Journal of Chemometrics, Volume 2 (1988), pp. 211-228
[13] Continuous Univariate Distributions, 2, Wiley, New York, 1995 | Zbl
[14] A Generic Algorithm for Reducing Bias in Parametric Estimation, Journal of Chemometrics, Volume 4 (2010), pp. 1097-1112 | Zbl
[15] The Degrees of Freedom of Partial Least Squares Regression, Journal of the American Statistical Association, Volume 106 (2011) no. 494, pp. 697-705 | Zbl
[16] Model selection for partial least squares regression, Chemometrics and Intelligent Laboratory Systems, Volume 64 (2002), pp. 79-89
[17] Comparaison de variantes de régressions logistiques PLS et de régression PLS sur variables qualitatives : application aux donnés d’allélotypage, Journal de la Société Française De Statistique, Volume 151 (2010) no. 2, pp. 1-18 | Numdam | Zbl
[18] Generalized Linear Models, Chapman & Hall/CRC, Boca Raton, 1995 | Zbl
[19] Natural exponential families with quadratic variance functions, The Annals of Statistics, Volume 10 (1982), pp. 65-80 | Zbl
[20] Comparison of prediction methods for multicollinear data, Communications in Statistics – Simulation and Computation, Volume 14 (1985), pp. 545-576 | Zbl
[21] Generalized Linear Models, Journal of the Royal Statistical Society. Series A (General), Volume 135 (1972) no. 3, pp. 370-384
[22] A general class of zero-or-one inflated beta regression models., Computational Statistics & Data Analysis, Volume 56 (2012) no. 1, pp. 1609-1623 | Zbl
[23] Towards real-time metabolic profiling of a biopsy specimen during a surgical operation by 1H high resolution magic angle spinning nuclear magnetic resonance : a case report., Journal of Medical Case Reports, Volume 6 (2012) no. 1
[24] Improved Estimators for a General Class of Beta Regression Models, Computational Statistics & Data Analysis, Volume 54 (2010) no. 2, pp. 348-366 | Zbl
[25] La régression PLS : Théorie et Pratique, Technip, Paris, 1998 | Zbl
[26] Estimation of principal component and related models by iterative least squares, Multivariate Analysis (Krishnaiah, P.R., ed.), Academic Press, New York, 1966, pp. 391-420 | Zbl
[27] PLS-regression : a basic tool of Chemometrics, Chemometrics and Intelligent Laboratory Systems, Volume 58 (2001), pp. 109-130