Adaptive Metric Kernel Regression

Cyril Goutte; Jan Larsen

doi:10.1109/NNSP.1998.710648

Adaptive Metric Kernel Regression

Cyril Goutte, Jan Larsen

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

494 Downloads (Pure)

Abstract

Kernel smoothing is a widely used nonparametric pattern recognition technique. By nature, it suffers from the curse of dimensionality and is usually difficult to apply to high input dimensions. In this paper, we propose an algorithm that adapts the input metric used in multivariate regression by minimising a cross-validation estimate of the generalisation error. This allows one to automatically adjust the importance of different dimensions. The improvement in terms of modelling performance is illustrated on a variable selection task where the adaptive metric kernel clearly outperforms the standard approach

Original language	English
Title of host publication	Neural Networks for Signal Processing VIII, 1998. Proceedings of the 1998 IEEE Signal Processing Society Workshop
Place of Publication	Piscataway
Publisher	IEEE
Publication date	1998
Pages	184-193
ISBN (Print)	0-7803-5060-X
DOIs	https://doi.org/10.1109/NNSP.1998.710648
Publication status	Published - 1998
Event	NNSP´98, Neural Networks for Signal Processing VIII - Cambridge, U.K. Duration: 1 Jan 1998 → …

Conference

Conference	NNSP´98, Neural Networks for Signal Processing VIII
City	Cambridge, U.K.
Period	01/01/1998 → …

Bibliographical note

Copyright: 1998 IEEE. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE

Access to Document

10.1109/NNSP.1998.710648

GoutteFinal published version, 587 KB

OpenUrl availability

Full text

Cite this

@inproceedings{9474198d51864f8da0bc0119261313b4,

title = "Adaptive Metric Kernel Regression",

abstract = "Kernel smoothing is a widely used nonparametric pattern recognition technique. By nature, it suffers from the curse of dimensionality and is usually difficult to apply to high input dimensions. In this paper, we propose an algorithm that adapts the input metric used in multivariate regression by minimising a cross-validation estimate of the generalisation error. This allows one to automatically adjust the importance of different dimensions. The improvement in terms of modelling performance is illustrated on a variable selection task where the adaptive metric kernel clearly outperforms the standard approach",

author = "Cyril Goutte and Jan Larsen",

note = "Copyright: 1998 IEEE. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE; NNSP´98, Neural Networks for Signal Processing VIII ; Conference date: 01-01-1998",

year = "1998",

doi = "10.1109/NNSP.1998.710648",

language = "English",

isbn = "0-7803-5060-X",

pages = "184--193",

booktitle = "Neural Networks for Signal Processing VIII, 1998. Proceedings of the 1998 IEEE Signal Processing Society Workshop",

publisher = "IEEE",

address = "United States",

}

TY - GEN

T1 - Adaptive Metric Kernel Regression

AU - Goutte, Cyril

AU - Larsen, Jan

N1 - Copyright: 1998 IEEE. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE

PY - 1998

Y1 - 1998

N2 - Kernel smoothing is a widely used nonparametric pattern recognition technique. By nature, it suffers from the curse of dimensionality and is usually difficult to apply to high input dimensions. In this paper, we propose an algorithm that adapts the input metric used in multivariate regression by minimising a cross-validation estimate of the generalisation error. This allows one to automatically adjust the importance of different dimensions. The improvement in terms of modelling performance is illustrated on a variable selection task where the adaptive metric kernel clearly outperforms the standard approach

AB - Kernel smoothing is a widely used nonparametric pattern recognition technique. By nature, it suffers from the curse of dimensionality and is usually difficult to apply to high input dimensions. In this paper, we propose an algorithm that adapts the input metric used in multivariate regression by minimising a cross-validation estimate of the generalisation error. This allows one to automatically adjust the importance of different dimensions. The improvement in terms of modelling performance is illustrated on a variable selection task where the adaptive metric kernel clearly outperforms the standard approach

U2 - 10.1109/NNSP.1998.710648

DO - 10.1109/NNSP.1998.710648

M3 - Article in proceedings

SN - 0-7803-5060-X

SP - 184

EP - 193

BT - Neural Networks for Signal Processing VIII, 1998. Proceedings of the 1998 IEEE Signal Processing Society Workshop

PB - IEEE

CY - Piscataway

T2 - NNSP´98, Neural Networks for Signal Processing VIII

Y2 - 1 January 1998

ER -

Adaptive Metric Kernel Regression

Abstract

Conference

Bibliographical note

Access to Document

OpenUrl availability

Fingerprint

Cite this