Model-based overlapping clustering

Arindam Banerjee; Chase Krumpelman; Joydeep Ghosh; Sugato Basu; Raymond J. Mooney

doi:10.1145/1081870.1081932

Model-based overlapping clustering

Arindam Banerjee, Chase Krumpelman, Joydeep Ghosh, Sugato Basu, Raymond J. Mooney

Research output: Contribution to conference › Paper › peer-review

Abstract

While the vast majority of clustering algorithms are partitional, many real world datasets have inherently overlapping clusters. Several approaches to finding overlapping clusters have come from work on analysis of biological datasets. In this paper, we interpret an overlapping clustering model proposed by Segal et al. [23] as a generalization of Gaussian mixture models, and we extend it to an overlapping clustering model based on mixtures of any regular exponential family distribution and the corresponding Bregman divergence. We provide the necessary algorithm modifications for this extension, and present results on synthetic data as well as subsets of 20-Newsgroups and EachMovie datasets.

Original language	English (US)
Pages	532-537
Number of pages	6
DOIs	https://doi.org/10.1145/1081870.1081932
State	Published - 2005
Externally published	Yes
Event	KDD-2005: 11th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining - Chicago, IL, United States Duration: Aug 21 2005 → Aug 24 2005

Other

Other	KDD-2005: 11th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining
Country/Territory	United States
City	Chicago, IL
Period	8/21/05 → 8/24/05

Keywords

Bregman divergences
Exponential model
Graphical model
High-dimensional clustering
Overlapping clustering

ASJC Scopus subject areas

Software
Information Systems

Online availability

10.1145/1081870.1081932

Library availability

Discover UIUC Full Text

Cite this

@conference{4a8eba69468f44bb83878ca0dabe7a46,

title = "Model-based overlapping clustering",

abstract = "While the vast majority of clustering algorithms are partitional, many real world datasets have inherently overlapping clusters. Several approaches to finding overlapping clusters have come from work on analysis of biological datasets. In this paper, we interpret an overlapping clustering model proposed by Segal et al. [23] as a generalization of Gaussian mixture models, and we extend it to an overlapping clustering model based on mixtures of any regular exponential family distribution and the corresponding Bregman divergence. We provide the necessary algorithm modifications for this extension, and present results on synthetic data as well as subsets of 20-Newsgroups and EachMovie datasets.",

keywords = "Bregman divergences, Exponential model, Graphical model, High-dimensional clustering, Overlapping clustering",

author = "Arindam Banerjee and Chase Krumpelman and Joydeep Ghosh and Sugato Basu and Mooney, {Raymond J.}",

note = "Copyright: Copyright 2008 Elsevier B.V., All rights reserved.; KDD-2005: 11th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining ; Conference date: 21-08-2005 Through 24-08-2005",

year = "2005",

doi = "10.1145/1081870.1081932",

language = "English (US)",

pages = "532--537",

}

TY - CONF

T1 - Model-based overlapping clustering

AU - Banerjee, Arindam

AU - Krumpelman, Chase

AU - Ghosh, Joydeep

AU - Basu, Sugato

AU - Mooney, Raymond J.

PY - 2005

Y1 - 2005

N2 - While the vast majority of clustering algorithms are partitional, many real world datasets have inherently overlapping clusters. Several approaches to finding overlapping clusters have come from work on analysis of biological datasets. In this paper, we interpret an overlapping clustering model proposed by Segal et al. [23] as a generalization of Gaussian mixture models, and we extend it to an overlapping clustering model based on mixtures of any regular exponential family distribution and the corresponding Bregman divergence. We provide the necessary algorithm modifications for this extension, and present results on synthetic data as well as subsets of 20-Newsgroups and EachMovie datasets.

AB - While the vast majority of clustering algorithms are partitional, many real world datasets have inherently overlapping clusters. Several approaches to finding overlapping clusters have come from work on analysis of biological datasets. In this paper, we interpret an overlapping clustering model proposed by Segal et al. [23] as a generalization of Gaussian mixture models, and we extend it to an overlapping clustering model based on mixtures of any regular exponential family distribution and the corresponding Bregman divergence. We provide the necessary algorithm modifications for this extension, and present results on synthetic data as well as subsets of 20-Newsgroups and EachMovie datasets.

KW - Bregman divergences

KW - Exponential model

KW - Graphical model

KW - High-dimensional clustering

KW - Overlapping clustering

UR - http://www.scopus.com/inward/record.url?scp=32344441237&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=32344441237&partnerID=8YFLogxK

U2 - 10.1145/1081870.1081932

DO - 10.1145/1081870.1081932

M3 - Paper

AN - SCOPUS:32344441237

SP - 532

EP - 537

T2 - KDD-2005: 11th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining

Y2 - 21 August 2005 through 24 August 2005

ER -

Model-based overlapping clustering

Abstract

Other

Keywords

ASJC Scopus subject areas

Online availability

Library availability

Related links

Fingerprint

Cite this