Confidence Intervals for Adaptive Regression Estimation - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Reports Year : 1998

Confidence Intervals for Adaptive Regression Estimation

Oleg Lepski
  • Function : Author

Abstract

The problem of adaptive estimation of the regression function f from noisy observation is concerned. A wavelet adaptive estimator of unknown function with the confidence interval for the L_2-error are provided. We show that if f belongs to a Sobolev class, the proposed estimate and the associated confidence interval are minimax.
Fichier principal
Vignette du fichier
RR-3580.pdf (343.16 Ko) Télécharger le fichier

Dates and versions

inria-00073101 , version 1 (24-05-2006)

Identifiers

  • HAL Id : inria-00073101 , version 1

Cite

Anatoli B. Juditsky, Oleg Lepski. Confidence Intervals for Adaptive Regression Estimation. RR-3580, INRIA. 1998. ⟨inria-00073101⟩
85 View
152 Download

Share

Gmail Facebook Twitter LinkedIn More