Please use this identifier to cite or link to this item: https://doi.org/10.1162/08997660151134343
DC FieldValue
dc.titlePredictive approaches for choosing hyperparameters in gaussian processes
dc.contributor.authorSundararajan, S.
dc.contributor.authorKeerthi, S.S.
dc.date.accessioned2014-06-17T06:31:30Z
dc.date.available2014-06-17T06:31:30Z
dc.date.issued2001-05
dc.identifier.citationSundararajan, S., Keerthi, S.S. (2001-05). Predictive approaches for choosing hyperparameters in gaussian processes. Neural Computation 13 (5) : 1103-1118. ScholarBank@NUS Repository. https://doi.org/10.1162/08997660151134343
dc.identifier.issn08997667
dc.identifier.urihttp://scholarbank.nus.edu.sg/handle/10635/61140
dc.description.abstractGaussian 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 (CV) 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.
dc.description.urihttp://libproxy1.nus.edu.sg/login?url=http://dx.doi.org/10.1162/08997660151134343
dc.sourceScopus
dc.typeArticle
dc.contributor.departmentMECHANICAL ENGINEERING
dc.description.doi10.1162/08997660151134343
dc.description.sourcetitleNeural Computation
dc.description.volume13
dc.description.issue5
dc.description.page1103-1118
dc.identifier.isiut000168383000007
Appears in Collections:Staff Publications

Show simple item record
Files in This Item:
There are no files associated with this item.

SCOPUSTM   
Citations

67
checked on Dec 2, 2020

WEB OF SCIENCETM
Citations

61
checked on Dec 2, 2020

Page view(s)

52
checked on Nov 30, 2020

Google ScholarTM

Check

Altmetric


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.