Grid-Based Hierarchical Spatial Clustering Algorithm in Presence of Obstacle and Constraints

Clustering of spatial data in the presence of obstacles and constraints has the very strong practical value, and becomes to an important research issue. Most of the existing spatial clustering algorithm in presence of obstacles and constraints can't cluster with irregular obstacles and the vera...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:International Conference on Internet Computing for Science and Engineering S. 383 - 388
Hauptverfasser: Yue Yang, Jian-pei Zhang, Jing Yang
Format: Tagungsbericht
Sprache:Englisch
Veröffentlicht: IEEE 01.01.2008
Schlagworte:
Clustering algorithms
Costs
Data engineering
Data mining
Educational institutions
Graphics
grid
Grid computing
hierarchical
Internet
obstacles
Partitioning algorithms
Shape
spatial clustering
ISSN:2330-9857
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Clustering of spatial data in the presence of obstacles and constraints has the very strong practical value, and becomes to an important research issue. Most of the existing spatial clustering algorithm in presence of obstacles and constraints can't cluster with irregular obstacles and the veracity of clustering result is affected. The algorithm complexity is affected by complexity of computing obstacle-distance. Grid-based hierarchical spatial clustering algorithm which is abbreviated as GSHCOC is proposed. The advantage of grid-based clustering algorithm is inherited. The obstacle-grid is defined and the algorithm processes arbitrary shape obstacle and finds arbitrary shape clusters efficiently. Meanwhile, the hierarchical strategy is used to reduce the complexity of clustering in presence of obstacles and constraints and the operation efficiency of algorithm is improved. The results of experiment show that GSHCOC algorithm can process spatial clustering in presence of obstacles and constraints and has higher clustering quality and better performance.
ISSN:2330-9857
DOI:10.1109/ICICSE.2008.18