Details
| Originalsprache | Englisch |
|---|---|
| Titel des Sammelwerks | Proceedings - 2015 IEEE Symposium Series on Computational Intelligence, SSCI 2015 |
| Herausgeber (Verlag) | Institute of Electrical and Electronics Engineers Inc. |
| Seiten | 749-755 |
| Seitenumfang | 7 |
| ISBN (elektronisch) | 9781479975600 |
| Publikationsstatus | Veröffentlicht - 2015 |
| Extern publiziert | Ja |
| Veranstaltung | IEEE Symposium Series on Computational Intelligence, SSCI 2015 - Cape Town, Südafrika Dauer: 8 Dez. 2015 → 10 Dez. 2015 |
Publikationsreihe
| Name | Proceedings - 2015 IEEE Symposium Series on Computational Intelligence, SSCI 2015 |
|---|
Abstract
DBSCAN is one of the most common density-based clustering algorithms. While multiple works tried to present an appropriate estimate for needed parameters we propose an alternating optimization algorithm, which finds a locally optimal parameter combination. The algorithm is based on the combination of two hierarchical versions of DBSCAN, which can be generated by fixing one parameter and iterating through possible values of the second parameter. Due to monotonicity of the neighborhood sets and the core-condition, successive levels of the hierarchy can efficiently be computed. An local optimal parameter combination can be determined using internal cluster validation measures. In this work we are comparing the measures edge-correlation and silhouette coefficient. For the latter we propose a density-based interpretation and show a respective computational efficient estimate to detect non-convex clusters produced by DBSCAN. Our results show, that the algorithm can automatically detect a good DBSCAN clustering on a variety of cluster scenarios.
ASJC Scopus Sachgebiete
- Informatik (insg.)
- Artificial intelligence
Zitieren
- Standard
- Harvard
- Apa
- Vancouver
- BibTex
- RIS
Proceedings - 2015 IEEE Symposium Series on Computational Intelligence, SSCI 2015. Institute of Electrical and Electronics Engineers Inc., 2015. S. 749-755 7376687 (Proceedings - 2015 IEEE Symposium Series on Computational Intelligence, SSCI 2015).
Publikation: Beitrag in Buch/Bericht/Sammelwerk/Konferenzband › Aufsatz in Konferenzband › Forschung › Peer-Review
}
TY - GEN
T1 - An alternating optimization approach based on hierarchical adaptations of DBSCAN
AU - Dockhorn, Alexander
AU - Braune, Christian
AU - Kruse, Rudolf
N1 - Publisher Copyright: © 2015 IEEE.
PY - 2015
Y1 - 2015
N2 - DBSCAN is one of the most common density-based clustering algorithms. While multiple works tried to present an appropriate estimate for needed parameters we propose an alternating optimization algorithm, which finds a locally optimal parameter combination. The algorithm is based on the combination of two hierarchical versions of DBSCAN, which can be generated by fixing one parameter and iterating through possible values of the second parameter. Due to monotonicity of the neighborhood sets and the core-condition, successive levels of the hierarchy can efficiently be computed. An local optimal parameter combination can be determined using internal cluster validation measures. In this work we are comparing the measures edge-correlation and silhouette coefficient. For the latter we propose a density-based interpretation and show a respective computational efficient estimate to detect non-convex clusters produced by DBSCAN. Our results show, that the algorithm can automatically detect a good DBSCAN clustering on a variety of cluster scenarios.
AB - DBSCAN is one of the most common density-based clustering algorithms. While multiple works tried to present an appropriate estimate for needed parameters we propose an alternating optimization algorithm, which finds a locally optimal parameter combination. The algorithm is based on the combination of two hierarchical versions of DBSCAN, which can be generated by fixing one parameter and iterating through possible values of the second parameter. Due to monotonicity of the neighborhood sets and the core-condition, successive levels of the hierarchy can efficiently be computed. An local optimal parameter combination can be determined using internal cluster validation measures. In this work we are comparing the measures edge-correlation and silhouette coefficient. For the latter we propose a density-based interpretation and show a respective computational efficient estimate to detect non-convex clusters produced by DBSCAN. Our results show, that the algorithm can automatically detect a good DBSCAN clustering on a variety of cluster scenarios.
UR - http://www.scopus.com/inward/record.url?scp=84964994147&partnerID=8YFLogxK
U2 - 10.1109/SSCI.2015.113
DO - 10.1109/SSCI.2015.113
M3 - Conference contribution
AN - SCOPUS:84964994147
T3 - Proceedings - 2015 IEEE Symposium Series on Computational Intelligence, SSCI 2015
SP - 749
EP - 755
BT - Proceedings - 2015 IEEE Symposium Series on Computational Intelligence, SSCI 2015
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - IEEE Symposium Series on Computational Intelligence, SSCI 2015
Y2 - 8 December 2015 through 10 December 2015
ER -