Predictive approaches for choosing hyperparameters in Gaussian processes

Sundararajan, S and Keerthi, SS (2001) Predictive approaches for choosing hyperparameters in Gaussian processes. In: Neural Computation, 13 (5). pp. 1103-1118.

PDF Predictive.pdf - Published Version Restricted to Registered users only Download (120kB) | Request a copy

Abstract

Gaussian processes are powerful regression models specified by parameterized mean and covariance functions. Standard approaches to choose these parameters (known by the name hyperparameters) are maximum likelihood and maximum a posteriori. In this article, we propose and investigate predictive approaches based on Geisser's predictive sample reuse (PSR) methodology and the related Stone's cross-validation ICV) methodology. More specifically, we derive results for Geisser's surrogate predictive probability (GPP), Geisser's predictive mean square error (GPE), and the standard CV error and make a comparative study. Within an approximation we arrive at the generalized cross-validation (GCV) and establish its relationship with the GPP and GPE approaches. These approaches are tested on a number of problems. Experimental results show that these approaches are strongly competitive with the existing approaches.

Item Type:	Journal Article
Publication:	Neural Computation
Publisher:	MIT Press
Additional Information:	Copyright of this article belongs to M I T Press.
Keywords:	Cross-Validation
Department/Centre:	Division of Electrical Sciences > Computer Science & Automation
Date Deposited:	10 Feb 2010 07:20
Last Modified:	19 Sep 2010 04:56
URI:	http://eprints.iisc.ac.in/id/eprint/17309

Actions (login required)

View Item