Learning curves for Gaussian processes models: fluctuations and universality

Dorthe Malzahn; Manfred Opper

doi:10.1007/3-540-44668-0_39

Learning curves for Gaussian processes models: fluctuations and universality

Dorthe Malzahn, Manfred Opper

Research output: Chapter in Book/Published conference output › Chapter

Abstract

Based on a statistical mechanics approach, we develop a method for approximately computing average case learning curves and their sample fluctuations for Gaussian process regression models. We give examples for the Wiener process and show that universal relations (that are independent of the input distribution) between error measures can be derived.

Original language	English
Title of host publication	Artificial Neural Networks — ICANN 2001
Editors	G. Dorffner, H. Bischof, K. Hornik
Place of Publication	Berlin Heidelberg
Publisher	Springer
Pages	271-276
Number of pages	6
Volume	2130
ISBN (Print)	9783540424864
DOIs	https://doi.org/10.1007/3-540-44668-0_39
Publication status	Published - 1 Jan 2001
Event	Artificial Neural Networks 2001 - Vienna, Austria Duration: 21 Aug 2001 → 25 Aug 2001

Publication series

Name	Lecture Notes in Computer Science
Publisher	Springer

Conference

Conference	Artificial Neural Networks 2001
Abbreviated title	ICANN 2001
Country/Territory	Austria
City	Vienna
Period	21/08/01 → 25/08/01

Bibliographical note

The original publication is available at www.springerlink.com

Keywords

statistical
mechanics
computing average
learning curves
sample fluctuations
Gaussian process regression
Wiener process
universal relations
error measures

Access to Document

10.1007/3-540-44668-0_39

NCRG_2001_017.pdf
The original publication is available at www.springerlink.com
Accepted author manuscript, 176 KB

Cite this

@inbook{d4e0cc4f33b94e1dbe59b32e7b62c766,

title = "Learning curves for Gaussian processes models: fluctuations and universality",

abstract = "Based on a statistical mechanics approach, we develop a method for approximately computing average case learning curves and their sample fluctuations for Gaussian process regression models. We give examples for the Wiener process and show that universal relations (that are independent of the input distribution) between error measures can be derived.",

keywords = "statistical, mechanics, computing average, learning curves, sample fluctuations, Gaussian process regression, Wiener process, universal relations, error measures",

author = "Dorthe Malzahn and Manfred Opper",

note = "The original publication is available at www.springerlink.com; Artificial Neural Networks 2001, ICANN 2001 ; Conference date: 21-08-2001 Through 25-08-2001",

year = "2001",

month = jan,

day = "1",

doi = "10.1007/3-540-44668-0_39",

language = "English",

isbn = "9783540424864",

volume = "2130",

series = "Lecture Notes in Computer Science",

publisher = "Springer",

pages = "271--276",

editor = "G. Dorffner and H. Bischof and K. Hornik",

booktitle = "Artificial Neural Networks — ICANN 2001",

address = "Germany",

}

Malzahn, D & Opper, M 2001, Learning curves for Gaussian processes models: fluctuations and universality. in G Dorffner, H Bischof & K Hornik (eds), Artificial Neural Networks — ICANN 2001. vol. 2130, Lecture Notes in Computer Science, Springer, Berlin Heidelberg, pp. 271-276, Artificial Neural Networks 2001, Vienna, Austria, 21/08/01. https://doi.org/10.1007/3-540-44668-0_39

TY - CHAP

T1 - Learning curves for Gaussian processes models: fluctuations and universality

AU - Malzahn, Dorthe

AU - Opper, Manfred

N1 - The original publication is available at www.springerlink.com

PY - 2001/1/1

Y1 - 2001/1/1

N2 - Based on a statistical mechanics approach, we develop a method for approximately computing average case learning curves and their sample fluctuations for Gaussian process regression models. We give examples for the Wiener process and show that universal relations (that are independent of the input distribution) between error measures can be derived.

AB - Based on a statistical mechanics approach, we develop a method for approximately computing average case learning curves and their sample fluctuations for Gaussian process regression models. We give examples for the Wiener process and show that universal relations (that are independent of the input distribution) between error measures can be derived.

KW - statistical

KW - mechanics

KW - computing average

KW - learning curves

KW - sample fluctuations

KW - Gaussian process regression

KW - Wiener process

KW - universal relations

KW - error measures

UR - http://www.springerlink.com/content/70hdlkvp30u92dev/

U2 - 10.1007/3-540-44668-0_39

DO - 10.1007/3-540-44668-0_39

M3 - Chapter

SN - 9783540424864

VL - 2130

T3 - Lecture Notes in Computer Science

SP - 271

EP - 276

BT - Artificial Neural Networks — ICANN 2001

A2 - Dorffner, G.

A2 - Bischof, H.

A2 - Hornik, K.

PB - Springer

CY - Berlin Heidelberg

T2 - Artificial Neural Networks 2001

Y2 - 21 August 2001 through 25 August 2001

ER -

Learning curves for Gaussian processes models: fluctuations and universality

Abstract

Publication series

Conference

Bibliographical note

Keywords

Access to Document

Other files and links

Fingerprint

Cite this