Clustering based on density estimation with sparse grids

Benjamin Peherstorfer, Dirk Pflüger, Hans Joachim Bungartz

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

We present a density-based clustering method. The clusters are determined by splitting a similarity graph of the data into connected components. The splitting is accomplished by removing vertices of the graph at which an estimated density function of the data evaluates to values below a threshold. The density function is approximated on a sparse grid in order to make the method feasible in higher-dimensional settings and scalable in the number of data points. With benchmark examples we show that our method is competitive with other modern clustering methods. Furthermore, we consider a real-world example where we cluster nodes of a finite element model of a Chevrolet pick-up truck with respect to the displacements of the nodes during a frontal crash.

Original languageEnglish (US)
Title of host publicationKI 2012
Subtitle of host publicationAdvances in Artificial Intelligence - 35th Annual German Conference on AI, Proceedings
Pages131-142
Number of pages12
DOIs
StatePublished - Nov 6 2012
Event35th Annual German Conference on Artificial Intelligence, KI 2012 - Saarbrucken, Germany
Duration: Sep 24 2012Sep 27 2012

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume7526 LNAI
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other35th Annual German Conference on Artificial Intelligence, KI 2012
CountryGermany
CitySaarbrucken
Period9/24/129/27/12

Fingerprint

Sparse Grids
Density Estimation
Probability density function
Clustering
Clustering Methods
Density Function
Trucks
Crash
Graph in graph theory
Vertex of a graph
Connected Components
Finite Element Model
High-dimensional
Benchmark
Evaluate

Keywords

  • clustering
  • density estimation
  • sparse grids

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Peherstorfer, B., Pflüger, D., & Bungartz, H. J. (2012). Clustering based on density estimation with sparse grids. In KI 2012: Advances in Artificial Intelligence - 35th Annual German Conference on AI, Proceedings (pp. 131-142). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7526 LNAI). https://doi.org/10.1007/978-3-642-33347-7_12

Clustering based on density estimation with sparse grids. / Peherstorfer, Benjamin; Pflüger, Dirk; Bungartz, Hans Joachim.

KI 2012: Advances in Artificial Intelligence - 35th Annual German Conference on AI, Proceedings. 2012. p. 131-142 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7526 LNAI).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Peherstorfer, B, Pflüger, D & Bungartz, HJ 2012, Clustering based on density estimation with sparse grids. in KI 2012: Advances in Artificial Intelligence - 35th Annual German Conference on AI, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 7526 LNAI, pp. 131-142, 35th Annual German Conference on Artificial Intelligence, KI 2012, Saarbrucken, Germany, 9/24/12. https://doi.org/10.1007/978-3-642-33347-7_12
Peherstorfer B, Pflüger D, Bungartz HJ. Clustering based on density estimation with sparse grids. In KI 2012: Advances in Artificial Intelligence - 35th Annual German Conference on AI, Proceedings. 2012. p. 131-142. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-642-33347-7_12
Peherstorfer, Benjamin ; Pflüger, Dirk ; Bungartz, Hans Joachim. / Clustering based on density estimation with sparse grids. KI 2012: Advances in Artificial Intelligence - 35th Annual German Conference on AI, Proceedings. 2012. pp. 131-142 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{b9780ad041cd4bfab8c9a58d359ef891,
title = "Clustering based on density estimation with sparse grids",
abstract = "We present a density-based clustering method. The clusters are determined by splitting a similarity graph of the data into connected components. The splitting is accomplished by removing vertices of the graph at which an estimated density function of the data evaluates to values below a threshold. The density function is approximated on a sparse grid in order to make the method feasible in higher-dimensional settings and scalable in the number of data points. With benchmark examples we show that our method is competitive with other modern clustering methods. Furthermore, we consider a real-world example where we cluster nodes of a finite element model of a Chevrolet pick-up truck with respect to the displacements of the nodes during a frontal crash.",
keywords = "clustering, density estimation, sparse grids",
author = "Benjamin Peherstorfer and Dirk Pfl{\"u}ger and Bungartz, {Hans Joachim}",
year = "2012",
month = "11",
day = "6",
doi = "10.1007/978-3-642-33347-7_12",
language = "English (US)",
isbn = "9783642333460",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
pages = "131--142",
booktitle = "KI 2012",

}

TY - GEN

T1 - Clustering based on density estimation with sparse grids

AU - Peherstorfer, Benjamin

AU - Pflüger, Dirk

AU - Bungartz, Hans Joachim

PY - 2012/11/6

Y1 - 2012/11/6

N2 - We present a density-based clustering method. The clusters are determined by splitting a similarity graph of the data into connected components. The splitting is accomplished by removing vertices of the graph at which an estimated density function of the data evaluates to values below a threshold. The density function is approximated on a sparse grid in order to make the method feasible in higher-dimensional settings and scalable in the number of data points. With benchmark examples we show that our method is competitive with other modern clustering methods. Furthermore, we consider a real-world example where we cluster nodes of a finite element model of a Chevrolet pick-up truck with respect to the displacements of the nodes during a frontal crash.

AB - We present a density-based clustering method. The clusters are determined by splitting a similarity graph of the data into connected components. The splitting is accomplished by removing vertices of the graph at which an estimated density function of the data evaluates to values below a threshold. The density function is approximated on a sparse grid in order to make the method feasible in higher-dimensional settings and scalable in the number of data points. With benchmark examples we show that our method is competitive with other modern clustering methods. Furthermore, we consider a real-world example where we cluster nodes of a finite element model of a Chevrolet pick-up truck with respect to the displacements of the nodes during a frontal crash.

KW - clustering

KW - density estimation

KW - sparse grids

UR - http://www.scopus.com/inward/record.url?scp=84868236159&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84868236159&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-33347-7_12

DO - 10.1007/978-3-642-33347-7_12

M3 - Conference contribution

AN - SCOPUS:84868236159

SN - 9783642333460

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

SP - 131

EP - 142

BT - KI 2012

ER -