Correlation clustering with stochastic labellings

Nicola Rebagliati, S. Rota Bulo, M. Pelillo

Research output: Chapter in Book/Report/Conference proceedingConference article in proceedingsScientificpeer-review

2 Citations (Scopus)

Abstract

Correlation clustering is the problem of finding a crisp partition of the vertices of a correlation graph in such a way as to minimize the disagreements in the cluster assignments. In this paper, we discuss a relaxation to the original problem setting which allows probabilistic assignments of vertices to labels. By so doing, overlapping clusters can be captured. We also show that a known optimization heuristic can be applied to the problem formulation, but with the automatic selection of the number of classes. Additionally, we propose a simple way of building an ensemble of agreement functions sampled from a reproducing kernel Hilbert space, which allows to apply correlation clustering without the empirical estimation of pairwise correlation values.
Original languageEnglish
Title of host publicationSimilarity-Based Pattern Recognition
Subtitle of host publicationSIMBAD 2013
Place of PublicationBerlin, Heidelberg
PublisherSpringer
Pages120-133
ISBN (Electronic)978-3-642-39140-8
ISBN (Print)978-3-642-39139-2
DOIs
Publication statusPublished - 2013
MoE publication typeNot Eligible
Event2nd International Workshop on Similarity-Based Pattern Analysis and Recognition, SIMBAD 2013 - York, United Kingdom
Duration: 3 Jul 20135 Jul 2013

Publication series

NameLecture Notes in Computer Science
PublisherSpringer
Volume7953
ISSN (Print)0302-9743

Conference

Conference2nd International Workshop on Similarity-Based Pattern Analysis and Recognition, SIMBAD 2013
Abbreviated titleSIMBAD 2013
CountryUnited Kingdom
CityYork
Period3/07/135/07/13

Fingerprint

Hilbert spaces
Labels

Keywords

  • Baum-Eagon inequality
  • correlation clustering
  • ensemble clustering
  • stochastic labelling

Cite this

Rebagliati, N., Rota Bulo, S., & Pelillo, M. (2013). Correlation clustering with stochastic labellings. In Similarity-Based Pattern Recognition: SIMBAD 2013 (pp. 120-133). Berlin, Heidelberg: Springer. Lecture Notes in Computer Science, Vol.. 7953 https://doi.org/10.1007/978-3-642-39140-8_8
Rebagliati, Nicola ; Rota Bulo, S. ; Pelillo, M. / Correlation clustering with stochastic labellings. Similarity-Based Pattern Recognition: SIMBAD 2013. Berlin, Heidelberg : Springer, 2013. pp. 120-133 (Lecture Notes in Computer Science, Vol. 7953 ).
@inproceedings{52b34e903ca540c995174795cc86d8c4,
title = "Correlation clustering with stochastic labellings",
abstract = "Correlation clustering is the problem of finding a crisp partition of the vertices of a correlation graph in such a way as to minimize the disagreements in the cluster assignments. In this paper, we discuss a relaxation to the original problem setting which allows probabilistic assignments of vertices to labels. By so doing, overlapping clusters can be captured. We also show that a known optimization heuristic can be applied to the problem formulation, but with the automatic selection of the number of classes. Additionally, we propose a simple way of building an ensemble of agreement functions sampled from a reproducing kernel Hilbert space, which allows to apply correlation clustering without the empirical estimation of pairwise correlation values.",
keywords = "Baum-Eagon inequality, correlation clustering, ensemble clustering, stochastic labelling",
author = "Nicola Rebagliati and {Rota Bulo}, S. and M. Pelillo",
year = "2013",
doi = "10.1007/978-3-642-39140-8_8",
language = "English",
isbn = "978-3-642-39139-2",
series = "Lecture Notes in Computer Science",
publisher = "Springer",
pages = "120--133",
booktitle = "Similarity-Based Pattern Recognition",
address = "Germany",

}

Rebagliati, N, Rota Bulo, S & Pelillo, M 2013, Correlation clustering with stochastic labellings. in Similarity-Based Pattern Recognition: SIMBAD 2013. Springer, Berlin, Heidelberg, Lecture Notes in Computer Science, vol. 7953 , pp. 120-133, 2nd International Workshop on Similarity-Based Pattern Analysis and Recognition, SIMBAD 2013, York, United Kingdom, 3/07/13. https://doi.org/10.1007/978-3-642-39140-8_8

Correlation clustering with stochastic labellings. / Rebagliati, Nicola; Rota Bulo, S.; Pelillo, M.

Similarity-Based Pattern Recognition: SIMBAD 2013. Berlin, Heidelberg : Springer, 2013. p. 120-133 (Lecture Notes in Computer Science, Vol. 7953 ).

Research output: Chapter in Book/Report/Conference proceedingConference article in proceedingsScientificpeer-review

TY - GEN

T1 - Correlation clustering with stochastic labellings

AU - Rebagliati, Nicola

AU - Rota Bulo, S.

AU - Pelillo, M.

PY - 2013

Y1 - 2013

N2 - Correlation clustering is the problem of finding a crisp partition of the vertices of a correlation graph in such a way as to minimize the disagreements in the cluster assignments. In this paper, we discuss a relaxation to the original problem setting which allows probabilistic assignments of vertices to labels. By so doing, overlapping clusters can be captured. We also show that a known optimization heuristic can be applied to the problem formulation, but with the automatic selection of the number of classes. Additionally, we propose a simple way of building an ensemble of agreement functions sampled from a reproducing kernel Hilbert space, which allows to apply correlation clustering without the empirical estimation of pairwise correlation values.

AB - Correlation clustering is the problem of finding a crisp partition of the vertices of a correlation graph in such a way as to minimize the disagreements in the cluster assignments. In this paper, we discuss a relaxation to the original problem setting which allows probabilistic assignments of vertices to labels. By so doing, overlapping clusters can be captured. We also show that a known optimization heuristic can be applied to the problem formulation, but with the automatic selection of the number of classes. Additionally, we propose a simple way of building an ensemble of agreement functions sampled from a reproducing kernel Hilbert space, which allows to apply correlation clustering without the empirical estimation of pairwise correlation values.

KW - Baum-Eagon inequality

KW - correlation clustering

KW - ensemble clustering

KW - stochastic labelling

U2 - 10.1007/978-3-642-39140-8_8

DO - 10.1007/978-3-642-39140-8_8

M3 - Conference article in proceedings

SN - 978-3-642-39139-2

T3 - Lecture Notes in Computer Science

SP - 120

EP - 133

BT - Similarity-Based Pattern Recognition

PB - Springer

CY - Berlin, Heidelberg

ER -

Rebagliati N, Rota Bulo S, Pelillo M. Correlation clustering with stochastic labellings. In Similarity-Based Pattern Recognition: SIMBAD 2013. Berlin, Heidelberg: Springer. 2013. p. 120-133. (Lecture Notes in Computer Science, Vol. 7953 ). https://doi.org/10.1007/978-3-642-39140-8_8