Modelling input varying correlations between multiple responses

Andrew Gordon Wilson; Zoubin Ghahramani

doi:10.1007/978-3-642-33486-3_64

Modelling input varying correlations between multiple responses

Andrew Gordon Wilson, Zoubin Ghahramani

Computer Science

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

Abstract

We introduced a generalised Wishart process (GWP) for modelling input dependent covariance matrices ∑(x), allowing one to model input varying correlations and uncertainties between multiple response variables. The GWP can naturally scale to thousands of response variables, as opposed to competing multivariate volatility models which are typically intractable for greater than 5 response variables. The GWP can also naturally capture a rich class of covariance dynamics - periodicity, Brownian motion, smoothness, ...- through a covariance kernel.

Original language	English (US)
Title of host publication	Machine Learning and Knowledge Discovery in Databases - European Conference, ECML PKDD 2012, Proceedings
Pages	858-861
Number of pages	4
Edition	PART 2
DOIs	https://doi.org/10.1007/978-3-642-33486-3_64
State	Published - 2012
Event	2012 European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases, ECML-PKDD 2012 - Bristol, United Kingdom Duration: Sep 24 2012 → Sep 28 2012

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Number	PART 2
Volume	7524 LNAI
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	2012 European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases, ECML-PKDD 2012
Country/Territory	United Kingdom
City	Bristol
Period	9/24/12 → 9/28/12

ASJC Scopus subject areas

Theoretical Computer Science
General Computer Science

Access to Document

10.1007/978-3-642-33486-3_64

Cite this

Wilson, A. G., & Ghahramani, Z. (2012). Modelling input varying correlations between multiple responses. In Machine Learning and Knowledge Discovery in Databases - European Conference, ECML PKDD 2012, Proceedings (PART 2 ed., pp. 858-861). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7524 LNAI, No. PART 2). https://doi.org/10.1007/978-3-642-33486-3_64

Modelling input varying correlations between multiple responses. / Wilson, Andrew Gordon; Ghahramani, Zoubin.
Machine Learning and Knowledge Discovery in Databases - European Conference, ECML PKDD 2012, Proceedings. PART 2. ed. 2012. p. 858-861 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7524 LNAI, No. PART 2).

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

Wilson, AG & Ghahramani, Z 2012, Modelling input varying correlations between multiple responses. in Machine Learning and Knowledge Discovery in Databases - European Conference, ECML PKDD 2012, Proceedings. PART 2 edn, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), no. PART 2, vol. 7524 LNAI, pp. 858-861, 2012 European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases, ECML-PKDD 2012, Bristol, United Kingdom, 9/24/12. https://doi.org/10.1007/978-3-642-33486-3_64

Wilson AG, Ghahramani Z. Modelling input varying correlations between multiple responses. In Machine Learning and Knowledge Discovery in Databases - European Conference, ECML PKDD 2012, Proceedings. PART 2 ed. 2012. p. 858-861. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); PART 2). doi: 10.1007/978-3-642-33486-3_64

Wilson, Andrew Gordon ; Ghahramani, Zoubin. / Modelling input varying correlations between multiple responses. Machine Learning and Knowledge Discovery in Databases - European Conference, ECML PKDD 2012, Proceedings. PART 2. ed. 2012. pp. 858-861 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); PART 2).

@inproceedings{9d727d2885644332bb2a9bb395af08a5,

title = "Modelling input varying correlations between multiple responses",

abstract = "We introduced a generalised Wishart process (GWP) for modelling input dependent covariance matrices ∑(x), allowing one to model input varying correlations and uncertainties between multiple response variables. The GWP can naturally scale to thousands of response variables, as opposed to competing multivariate volatility models which are typically intractable for greater than 5 response variables. The GWP can also naturally capture a rich class of covariance dynamics - periodicity, Brownian motion, smoothness, ...- through a covariance kernel.",

author = "Wilson, {Andrew Gordon} and Zoubin Ghahramani",

note = "Copyright: Copyright 2012 Elsevier B.V., All rights reserved.; 2012 European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases, ECML-PKDD 2012 ; Conference date: 24-09-2012 Through 28-09-2012",

year = "2012",

doi = "10.1007/978-3-642-33486-3_64",

language = "English (US)",

isbn = "9783642334856",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

number = "PART 2",

pages = "858--861",

booktitle = "Machine Learning and Knowledge Discovery in Databases - European Conference, ECML PKDD 2012, Proceedings",

edition = "PART 2",

}

TY - GEN

T1 - Modelling input varying correlations between multiple responses

AU - Wilson, Andrew Gordon

AU - Ghahramani, Zoubin

PY - 2012

Y1 - 2012

N2 - We introduced a generalised Wishart process (GWP) for modelling input dependent covariance matrices ∑(x), allowing one to model input varying correlations and uncertainties between multiple response variables. The GWP can naturally scale to thousands of response variables, as opposed to competing multivariate volatility models which are typically intractable for greater than 5 response variables. The GWP can also naturally capture a rich class of covariance dynamics - periodicity, Brownian motion, smoothness, ...- through a covariance kernel.

AB - We introduced a generalised Wishart process (GWP) for modelling input dependent covariance matrices ∑(x), allowing one to model input varying correlations and uncertainties between multiple response variables. The GWP can naturally scale to thousands of response variables, as opposed to competing multivariate volatility models which are typically intractable for greater than 5 response variables. The GWP can also naturally capture a rich class of covariance dynamics - periodicity, Brownian motion, smoothness, ...- through a covariance kernel.

UR - http://www.scopus.com/inward/record.url?scp=84866874999&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84866874999&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-33486-3_64

DO - 10.1007/978-3-642-33486-3_64

M3 - Conference contribution

AN - SCOPUS:84866874999

SN - 9783642334856

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 858

EP - 861

BT - Machine Learning and Knowledge Discovery in Databases - European Conference, ECML PKDD 2012, Proceedings

T2 - 2012 European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases, ECML-PKDD 2012

Y2 - 24 September 2012 through 28 September 2012

ER -

Modelling input varying correlations between multiple responses

Abstract

Publication series

Conference

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this