Universal clustering with regularization in probabilistic space

Vladimir Nikulin; Alex J. Smola

doi:10.1007/11510888_15

Universal clustering with regularization in probabilistic space

Vladimir Nikulin^*, Alex J. Smola

^*Corresponding author for this work

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

Abstract

We propose universal clustering in line with the concepts of universal estimation. In order to illustrate above model we introduce family of power loss functions in probabilistic space which is marginally linked to the Kullback-Leibler divergence. Above model proved to be effective in application to the synthetic data. Also, we consider large web-traffic dataset. The aim of the experiment is to explain and understand the way people interact with web sites. The paper proposes special regularization in order to ensure consistency of the corresponding clustering model.

Original language	English
Title of host publication	Machine Learning and Data Mining in Pattern Recognition - 4th International Conference, MLDM 2005, Proceedings
Publisher	Springer Verlag
Pages	142-152
Number of pages	11
ISBN (Print)	3540269231, 9783540269236
DOIs	https://doi.org/10.1007/11510888_15
Publication status	Published - 2005
Event	4th International Conference on Machine Learning and Data Mining in Pattern Recognition, MLDM 2005 - Leipzig, Germany Duration: 9 Jul 2005 → 11 Jul 2005

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	3587 LNAI
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	4th International Conference on Machine Learning and Data Mining in Pattern Recognition, MLDM 2005
Country/Territory	Germany
City	Leipzig
Period	9/07/05 → 11/07/05

Access to Document

10.1007/11510888_15

Cite this

Nikulin, V., & Smola, A. J. (2005). Universal clustering with regularization in probabilistic space. In Machine Learning and Data Mining in Pattern Recognition - 4th International Conference, MLDM 2005, Proceedings (pp. 142-152). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 3587 LNAI). Springer Verlag. https://doi.org/10.1007/11510888_15

Nikulin, Vladimir ; Smola, Alex J. / Universal clustering with regularization in probabilistic space. Machine Learning and Data Mining in Pattern Recognition - 4th International Conference, MLDM 2005, Proceedings. Springer Verlag, 2005. pp. 142-152 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{8f99ef19ce02499ea752f6c81e0ee930,

title = "Universal clustering with regularization in probabilistic space",

abstract = "We propose universal clustering in line with the concepts of universal estimation. In order to illustrate above model we introduce family of power loss functions in probabilistic space which is marginally linked to the Kullback-Leibler divergence. Above model proved to be effective in application to the synthetic data. Also, we consider large web-traffic dataset. The aim of the experiment is to explain and understand the way people interact with web sites. The paper proposes special regularization in order to ensure consistency of the corresponding clustering model.",

author = "Vladimir Nikulin and Smola, {Alex J.}",

year = "2005",

doi = "10.1007/11510888_15",

language = "English",

isbn = "3540269231",

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

publisher = "Springer Verlag",

pages = "142--152",

booktitle = "Machine Learning and Data Mining in Pattern Recognition - 4th International Conference, MLDM 2005, Proceedings",

address = "Germany",

note = "4th International Conference on Machine Learning and Data Mining in Pattern Recognition, MLDM 2005 ; Conference date: 09-07-2005 Through 11-07-2005",

}

Nikulin, V & Smola, AJ 2005, Universal clustering with regularization in probabilistic space. in Machine Learning and Data Mining in Pattern Recognition - 4th International Conference, MLDM 2005, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 3587 LNAI, Springer Verlag, pp. 142-152, 4th International Conference on Machine Learning and Data Mining in Pattern Recognition, MLDM 2005, Leipzig, Germany, 9/07/05. https://doi.org/10.1007/11510888_15

Universal clustering with regularization in probabilistic space. / Nikulin, Vladimir; Smola, Alex J.
Machine Learning and Data Mining in Pattern Recognition - 4th International Conference, MLDM 2005, Proceedings. Springer Verlag, 2005. p. 142-152 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 3587 LNAI).

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

TY - GEN

T1 - Universal clustering with regularization in probabilistic space

AU - Nikulin, Vladimir

AU - Smola, Alex J.

PY - 2005

Y1 - 2005

N2 - We propose universal clustering in line with the concepts of universal estimation. In order to illustrate above model we introduce family of power loss functions in probabilistic space which is marginally linked to the Kullback-Leibler divergence. Above model proved to be effective in application to the synthetic data. Also, we consider large web-traffic dataset. The aim of the experiment is to explain and understand the way people interact with web sites. The paper proposes special regularization in order to ensure consistency of the corresponding clustering model.

AB - We propose universal clustering in line with the concepts of universal estimation. In order to illustrate above model we introduce family of power loss functions in probabilistic space which is marginally linked to the Kullback-Leibler divergence. Above model proved to be effective in application to the synthetic data. Also, we consider large web-traffic dataset. The aim of the experiment is to explain and understand the way people interact with web sites. The paper proposes special regularization in order to ensure consistency of the corresponding clustering model.

UR - http://www.scopus.com/inward/record.url?scp=26944486169&partnerID=8YFLogxK

U2 - 10.1007/11510888_15

DO - 10.1007/11510888_15

M3 - Conference contribution

SN - 3540269231

SN - 9783540269236

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

SP - 142

EP - 152

BT - Machine Learning and Data Mining in Pattern Recognition - 4th International Conference, MLDM 2005, Proceedings

PB - Springer Verlag

T2 - 4th International Conference on Machine Learning and Data Mining in Pattern Recognition, MLDM 2005

Y2 - 9 July 2005 through 11 July 2005

ER -

Nikulin V, Smola AJ. Universal clustering with regularization in probabilistic space. In Machine Learning and Data Mining in Pattern Recognition - 4th International Conference, MLDM 2005, Proceedings. Springer Verlag. 2005. p. 142-152. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/11510888_15

Universal clustering with regularization in probabilistic space

Abstract

Publication series

Conference

Access to Document

Other files and links

Fingerprint

Cite this