Computational outlier detection methods in sliced inverse regression
SARACCO, Jérôme
Ecole Nationale Supérieure de Cognitique [ENSC]
Méthodes avancées d’apprentissage statistique et de contrôle [ASTRAL]
Ecole Nationale Supérieure de Cognitique [ENSC]
Méthodes avancées d’apprentissage statistique et de contrôle [ASTRAL]
SARACCO, Jérôme
Ecole Nationale Supérieure de Cognitique [ENSC]
Méthodes avancées d’apprentissage statistique et de contrôle [ASTRAL]
< Reduce
Ecole Nationale Supérieure de Cognitique [ENSC]
Méthodes avancées d’apprentissage statistique et de contrôle [ASTRAL]
Language
en
Chapitre d'ouvrage
This item was published in
Advances in Contemporary Statistics and Econometrics, Advances in Contemporary Statistics and Econometrics. 2021-06-15p. 101-122
Springer International Publishing
English Abstract
Sliced inverse regression (SIR) focuses on the relationship between a dependent variable y and a p-dimensional explanatory variable x in a semiparametric regression model in which the link relies on an index x β and link ...Read more >
Sliced inverse regression (SIR) focuses on the relationship between a dependent variable y and a p-dimensional explanatory variable x in a semiparametric regression model in which the link relies on an index x β and link function f. SIR allows to estimate the direction of β that forms the effective dimension reduction (EDR) space. Based on the estimated index, the link function f can then be nonparametrically estimated using kernel estimator. This two-step approach is sensitive to the presence of outliers in the data. The aim of this paper is to propose computational methods to detect outliers in that kind of single-index regression model. Three outlier detection methods are proposed and their numerical behaviors are illustrated on a simulated sample. To discriminate outliers from "normal" observations, they use IB (in-bags) or OOB (out-of-bags) prediction errors from subsampling or resampling approaches. These methods, implemented in R, are compared with each other in a simulation study. An application on a real data is also provided.Read less <
Origin
Hal imported