FUNDAMENTALNAYA I PRIKLADNAYA MATEMATIKA

(FUNDAMENTAL AND APPLIED MATHEMATICS)

2013, VOLUME 18, NUMBER 2, PAGES 53-65

Properties of the Bayesian parameter estimation of a regression based on Gaussian processes

A. A. Zaytsev
E. V. Burnaev
V. G. Spokoiny

Abstract

View as HTML     View as gif image

We consider the regression approach based on Gaussian processes and outline our theoretical results about the properties of the posterior distribution of the corresponding covariance function's parameter vector. We perform statistical experiments confirming that the obtained theoretical propositions are valid for a wide class of covariance functions commonly used in applied problems.

Location: http://mech.math.msu.su/~fpm/eng/k13/k132/k13204h.htm
Last modified: January 7, 2014