Comparer des méthodes

	Régression quantile robuste	Régression linéaire multiple robuste
Domaine	Statistique	Statistique
Famille	Regression model	Regression model
Année d'origine	1993–1997	1964–1980s
Auteur d'origine	Koenker & Bassett (1978); robust extensions by Machado (1993) and He (1997)	Peter J. Huber (M-estimators, 1964); extended by Rousseeuw, Yohai, and Maronna
Type	Robust semiparametric regression	Robust linear regression
Source fondatrice	Koenker, R. (2005). Quantile Regression. Cambridge University Press. ISBN: 978-0521608275	Huber, P. J. (1964). Robust estimation of a location parameter. Annals of Mathematical Statistics, 35(1), 73–101.
Alias	robust QR, outlier-resistant quantile regression, bounded-influence quantile regression, RQR	robust MLR, M-estimator regression, resistant multiple regression, robust OLS
Apparentées	6	6
Résumé	Robust Quantile Regression estimates conditional quantiles of a response variable while simultaneously downweighting the influence of outliers. By combining the asymmetric loss function of standard quantile regression with bounded-influence or M-estimation weights, it provides reliable quantile estimates even when data contain extreme observations or heavy-tailed error distributions.	Robust multiple linear regression estimates the linear relationship between a continuous outcome and several predictors while being resistant to outliers and violations of the normality assumption. Instead of minimising the sum of squared residuals, it uses a bounded loss function — most commonly Huber's or Tukey's bisquare — so that extreme observations receive limited influence on the estimated coefficients.