Maximum likelihood estimation and prediction error for a Matérn model on the circle - Archive ouverte HAL Access content directly
Preprints, Working Papers, ... Year :

Maximum likelihood estimation and prediction error for a Matérn model on the circle

Estimation par maximum de vraisemblance et erreur de prédiction pour un modèle de Matérn défini sur le cercle

(1, 2)
1
2

Abstract

This work considers Gaussian process interpolation with a periodized version of the Matérn covariance function (Stein, 1999, Section 6.7) with Fourier coefficients φ(α^2 + j^2)^(−ν−1/2). Convergence rates are studied for the joint maximum likelihood estimation of ν and φ when the data is sampled according to the model. The mean integrated squared error is also analyzed with fixed and estimated parameters, showing that maximum likelihood estimation yields asymptotically the same error as if the ground truth was known. Finally, the case where the observed function is a "deterministic" element of a continuous Sobolev space is also considered, suggesting that bounding assumptions on some parameters can lead to different estimates.
Fichier principal
Vignette du fichier
spetit_periodic_nu_paper.pdf (273.86 Ko) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-03778527 , version 1 (15-09-2022)
hal-03778527 , version 2 (07-11-2022)
hal-03778527 , version 3 (13-12-2022)
hal-03778527 , version 4 (30-01-2023)

Identifiers

Cite

Sébastien Petit. Maximum likelihood estimation and prediction error for a Matérn model on the circle. 2022. ⟨hal-03778527v3⟩
18 View
17 Download

Altmetric

Share

Gmail Facebook Twitter LinkedIn More