Presentation 2006-03-15
Statistical mechanics of Gaussian Process
Jun TSUZURUGI, Hideaki SAKAI,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) In this paper, we request the optimized condition and the analytic solution of the restoration error of Gaussian process by a statistical mechanical analysis. Concretely, the input space of the Gaussian process is assumed to be even intervals, and the periodic boundary condition is forced. As a result, the diagonalization of the kernel and the discussion on the Fourier space becomes possible. We could understand that restoration error was minimized when each hyperparameter of probabilistic model used in the restoration process coincided with that one used in formation process. In the optimization condition, the numerical solution and the restoration error of analytic solution were corresponded.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Gaussian process / Statistical mechanics / Optimization / Hyperparameter
Paper # NC2005-116
Date of Issue

Conference Information
Committee NC
Conference Date 2006/3/8(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Neurocomputing (NC)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Statistical mechanics of Gaussian Process
Sub Title (in English)
Keyword(1) Gaussian process
Keyword(2) Statistical mechanics
Keyword(3) Optimization
Keyword(4) Hyperparameter
1st Author's Name Jun TSUZURUGI
1st Author's Affiliation Okayama University of Science Faculty of Engineering:Graduate School of Informatics Kyoto University()
2nd Author's Name Hideaki SAKAI
2nd Author's Affiliation Graduate School of Informatics Kyoto University
Date 2006-03-15
Paper # NC2005-116
Volume (vol) vol.105
Number (no) 657
Page pp.pp.-
#Pages 4
Date of Issue