Estimation of the density of regression errors by pointwise model selection

Abstract : This paper presents two results: a density estimator and an estimator of regression error density. We first propose a density estimator constructed by model selection, which is adaptive for the quadratic risk at a given point. Then we apply this result to estimate the error density in an homoscedastic regression framework $Y_i=b(X_i) + \epsilon _i$, from which we observe a sample $(X_i,Y_i)$. Given an adaptive estimator $\widehat{b}$ of the regression function, we apply the density estimation procedure to the residuals $\widehat{\epsilon} _i = Y_i -\widehat{b} (X_i)$. We get an estimator of the density of $\epsilon _i$ whose rate of convergence for the quadratic pointwise risk is the maximum of two rates: the minimax rate we would get if the errors were directly observed and the minimax rate of convergence of $\widehat{b}$ for the quadratic integrated risk.
Document type :
Preprints, Working Papers, ...
Complete list of metadatas

Cited literature [16 references]  Display  Hide  Download
Contributor : Sandra Plancade <>
Submitted on : Thursday, February 26, 2009 - 12:15:55 PM
Last modification on : Friday, September 20, 2019 - 4:34:02 PM
Long-term archiving on: Friday, October 12, 2012 - 12:30:44 PM


Files produced by the author(s)


  • HAL Id : hal-00364334, version 1



Sandra Plancade. Estimation of the density of regression errors by pointwise model selection. 2009. ⟨hal-00364334⟩



Record views


Files downloads