A Significance Test for Covariates in Nonparametric Regression

Lavergne, Pascal, Maistre, Samuel and Patilea, Valentin (2015) A Significance Test for Covariates in Nonparametric Regression. Electronic Journal of Statistics, vol. 9. pp. 643-678.

This is the latest version of this item.

Download (370kB) | Preview
Official URL: http://tse-fr.eu/pub/29256


We consider testing the significance of a subset of covariates in a nonparamet- ric regression. These covariates can be continuous and/or discrete. We propose a new kernel-based test that smoothes only over the covariates appearing under the null hypothesis, so that the curse of dimensionality is mitigated. The test statistic is asymptotically pivotal and the rate of which the test detects local alternatives depends only on the dimension of the covariates under the null hy- pothesis. We show the validity of wild bootstrap for the test. In small samples, our test is competitive compared to existing procedures.

Item Type: Article
Language: English
Date: 2015
Refereed: Yes
Uncontrolled Keywords: Testing, Bootstrap, Kernel Smoothing, U−statistic
JEL codes: C14 - Semiparametric and Nonparametric Methods
C52 - Model Evaluation and Selection
Divisions: TSE-R (Toulouse)
Site: UT1
Date Deposited: 21 Sep 2015 13:08
Last Modified: 10 Apr 2018 09:59
OAI ID: oai:tse-fr.eu:29256
URI: http://publications.ut-capitole.fr/id/eprint/16880

Available Versions of this Item

Actions (login required)

View Item View Item


Downloads per month over past year