Bayesian clustering by dynamics

Marco Ramoni; Paola Sebastiani; Paul Cohen

doi:10.1023/A:1013635829250

Bayesian clustering by dynamics

Marco Ramoni, Paola Sebastiani, Paul Cohen

Sciences Technology and Arts, School of

Research output: Contribution to journal › Article › peer-review

116 Scopus citations

Abstract

This paper introduces a Bayesian method for clustering dynamic processes. The method models dynamics as Markov chains and then applies an agglomerative clustering procedure to discover the most probable set of clusters capturing different dynamics. To increase efficiency, the method uses an entropy-based heuristic search strategy. A controlled experiment suggests that the method is very accurate when applied to artificial time series in a broad range of conditions and, when applied to clustering sensor data from mobile robots, it produces clusters that are meaningful in the domain of application.

Original language	English (US)
Pages (from-to)	91-121
Number of pages	31
Journal	Machine Learning
Volume	47
Issue number	1
DOIs	https://doi.org/10.1023/A:1013635829250
State	Published - Apr 2002

Keywords

Bayesian learning
Clustering
Entropy
Heuristic search
Markov chains
Time series

ASJC Scopus subject areas

Software
Artificial Intelligence

Access to Document

10.1023/A:1013635829250

Fingerprint Dive into the research topics of 'Bayesian clustering by dynamics'. Together they form a unique fingerprint.

Cite this

@article{a2d1769f1c144f5a8f5fe94ee5fcf4c9,

title = "Bayesian clustering by dynamics",

abstract = "This paper introduces a Bayesian method for clustering dynamic processes. The method models dynamics as Markov chains and then applies an agglomerative clustering procedure to discover the most probable set of clusters capturing different dynamics. To increase efficiency, the method uses an entropy-based heuristic search strategy. A controlled experiment suggests that the method is very accurate when applied to artificial time series in a broad range of conditions and, when applied to clustering sensor data from mobile robots, it produces clusters that are meaningful in the domain of application.",

keywords = "Bayesian learning, Clustering, Entropy, Heuristic search, Markov chains, Time series",

author = "Marco Ramoni and Paola Sebastiani and Paul Cohen",

note = "Funding Information: We wish to thank Greg Ridgeway for his generous contribution and for making us available the source code of his program. We are also grateful to Doug Fisher and the anonymous referees for dramatically improving the quality of this paper. This research is supported by DARPA/AFOSR under contract(s) No(s) F49620-97-1-0485. The U.S. Government is authorized to reproduce and distribute reprints for governmental purposes notwithstanding any copyright notation hereon. The views and conclusions contained herein are those of the authors and should not be interpreted as necessarily representing the official policies or endorsements either expressed or implied, of DARPA/AFOSR or the U.S. Government.",

year = "2002",

month = apr,

doi = "10.1023/A:1013635829250",

language = "English (US)",

volume = "47",

pages = "91--121",

journal = "Machine Learning",

issn = "0885-6125",

publisher = "Springer Netherlands",

number = "1",

}

TY - JOUR

T1 - Bayesian clustering by dynamics

AU - Ramoni, Marco

AU - Sebastiani, Paola

AU - Cohen, Paul

N1 - Funding Information: We wish to thank Greg Ridgeway for his generous contribution and for making us available the source code of his program. We are also grateful to Doug Fisher and the anonymous referees for dramatically improving the quality of this paper. This research is supported by DARPA/AFOSR under contract(s) No(s) F49620-97-1-0485. The U.S. Government is authorized to reproduce and distribute reprints for governmental purposes notwithstanding any copyright notation hereon. The views and conclusions contained herein are those of the authors and should not be interpreted as necessarily representing the official policies or endorsements either expressed or implied, of DARPA/AFOSR or the U.S. Government.

PY - 2002/4

Y1 - 2002/4

N2 - This paper introduces a Bayesian method for clustering dynamic processes. The method models dynamics as Markov chains and then applies an agglomerative clustering procedure to discover the most probable set of clusters capturing different dynamics. To increase efficiency, the method uses an entropy-based heuristic search strategy. A controlled experiment suggests that the method is very accurate when applied to artificial time series in a broad range of conditions and, when applied to clustering sensor data from mobile robots, it produces clusters that are meaningful in the domain of application.

AB - This paper introduces a Bayesian method for clustering dynamic processes. The method models dynamics as Markov chains and then applies an agglomerative clustering procedure to discover the most probable set of clusters capturing different dynamics. To increase efficiency, the method uses an entropy-based heuristic search strategy. A controlled experiment suggests that the method is very accurate when applied to artificial time series in a broad range of conditions and, when applied to clustering sensor data from mobile robots, it produces clusters that are meaningful in the domain of application.

KW - Bayesian learning

KW - Clustering

KW - Entropy

KW - Heuristic search

KW - Markov chains

KW - Time series

UR - http://www.scopus.com/inward/record.url?scp=0036532775&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0036532775&partnerID=8YFLogxK

U2 - 10.1023/A:1013635829250

DO - 10.1023/A:1013635829250

M3 - Article

AN - SCOPUS:0036532775

VL - 47

SP - 91

EP - 121

JO - Machine Learning

JF - Machine Learning

SN - 0885-6125

IS - 1

ER -