Heteroscedastic Gaussian process regression

Quoc V. Le; Alex J. Smola; Stéphane Canu

doi:10.1145/1102351.1102413

Heteroscedastic Gaussian process regression

Quoc V. Le^*, Alex J. Smola, Stéphane Canu

^*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

148 Citations (Scopus)

Abstract

This paper presents an algorithm to estimate simultaneously both mean and variance of a non parametric regression problem. The key point is that we are able to estimate variance locally unlike standard Gaussian Process regression or SVMs. This means that our estimator adapts to the local noise. The problem is cast in the setting of maximum a posteriori estimation in exponential families. Unlike previous work, we obtain a convex optimization problem which can be solved via Newton's method.

Original language	English
Title of host publication	ICML 2005 - Proceedings of the 22nd International Conference on Machine Learning
Editors	L. Raedt, S. Wrobel
Pages	489-496
Number of pages	8
DOIs	https://doi.org/10.1145/1102351.1102413
Publication status	Published - 2005
Event	ICML 2005: 22nd International Conference on Machine Learning - Bonn, Germany Duration: 7 Aug 2005 → 11 Aug 2005

Publication series

Name	ICML 2005 - Proceedings of the 22nd International Conference on Machine Learning

Conference

Conference	ICML 2005: 22nd International Conference on Machine Learning
Country/Territory	Germany
City	Bonn
Period	7/08/05 → 11/08/05

Access to Document

10.1145/1102351.1102413

Cite this

@inproceedings{6d797637aeac475f90c6925632d64bce,

title = "Heteroscedastic Gaussian process regression",

abstract = "This paper presents an algorithm to estimate simultaneously both mean and variance of a non parametric regression problem. The key point is that we are able to estimate variance locally unlike standard Gaussian Process regression or SVMs. This means that our estimator adapts to the local noise. The problem is cast in the setting of maximum a posteriori estimation in exponential families. Unlike previous work, we obtain a convex optimization problem which can be solved via Newton's method.",

author = "Le, {Quoc V.} and Smola, {Alex J.} and St{\'e}phane Canu",

year = "2005",

doi = "10.1145/1102351.1102413",

language = "English",

isbn = "1595931805",

series = "ICML 2005 - Proceedings of the 22nd International Conference on Machine Learning",

pages = "489--496",

editor = "L. Raedt and S. Wrobel",

booktitle = "ICML 2005 - Proceedings of the 22nd International Conference on Machine Learning",

note = "ICML 2005: 22nd International Conference on Machine Learning ; Conference date: 07-08-2005 Through 11-08-2005",

}

Le, QV, Smola, AJ & Canu, S 2005, Heteroscedastic Gaussian process regression. in L Raedt & S Wrobel (eds), ICML 2005 - Proceedings of the 22nd International Conference on Machine Learning. ICML 2005 - Proceedings of the 22nd International Conference on Machine Learning, pp. 489-496, ICML 2005: 22nd International Conference on Machine Learning, Bonn, Germany, 7/08/05. https://doi.org/10.1145/1102351.1102413

Heteroscedastic Gaussian process regression. / Le, Quoc V.; Smola, Alex J.; Canu, Stéphane.
ICML 2005 - Proceedings of the 22nd International Conference on Machine Learning. ed. / L. Raedt; S. Wrobel. 2005. p. 489-496 (ICML 2005 - Proceedings of the 22nd International Conference on Machine Learning).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Heteroscedastic Gaussian process regression

AU - Le, Quoc V.

AU - Smola, Alex J.

AU - Canu, Stéphane

PY - 2005

Y1 - 2005

N2 - This paper presents an algorithm to estimate simultaneously both mean and variance of a non parametric regression problem. The key point is that we are able to estimate variance locally unlike standard Gaussian Process regression or SVMs. This means that our estimator adapts to the local noise. The problem is cast in the setting of maximum a posteriori estimation in exponential families. Unlike previous work, we obtain a convex optimization problem which can be solved via Newton's method.

AB - This paper presents an algorithm to estimate simultaneously both mean and variance of a non parametric regression problem. The key point is that we are able to estimate variance locally unlike standard Gaussian Process regression or SVMs. This means that our estimator adapts to the local noise. The problem is cast in the setting of maximum a posteriori estimation in exponential families. Unlike previous work, we obtain a convex optimization problem which can be solved via Newton's method.

UR - http://www.scopus.com/inward/record.url?scp=31844456216&partnerID=8YFLogxK

U2 - 10.1145/1102351.1102413

DO - 10.1145/1102351.1102413

M3 - Conference contribution

SN - 1595931805

T3 - ICML 2005 - Proceedings of the 22nd International Conference on Machine Learning

SP - 489

EP - 496

BT - ICML 2005 - Proceedings of the 22nd International Conference on Machine Learning

A2 - Raedt, L.

A2 - Wrobel, S.

T2 - ICML 2005: 22nd International Conference on Machine Learning

Y2 - 7 August 2005 through 11 August 2005

ER -

Heteroscedastic Gaussian process regression

Abstract

Publication series

Conference

Access to Document

Other files and links

Fingerprint

Cite this