Abstract | ||
---|---|---|
As a popular data mining tool, clustering focuses on revealing underlying patterns embedded in data. However, most existing clustering methods mainly deal with static data, which may not be suitable for analyzing large data in dynamic environments. To tackle this problem, this paper proposes an incremental clustering method based on the CFS, clustering by fast search and find of density peaks, to process large dynamic data. In the proposed method, multiple representatives are identified for each cluster to integrate new objects into previous clustering patterns at first. Then the convex hull theory is employed to modify the representatives accordingly. To further improve the generality and effectiveness, one-time cluster adjustment strategy is explored. Extensive experiments on several real-world image datasets demonstrate that the proposed method outperforms state-of-the-art methods for clustering large data. |
Year | Venue | Keywords |
---|---|---|
2017 | IEEE Global Conference on Signal and Information Processing | CFS,incremental clustering,objects assignment,clusters adjustment |
Field | DocType | ISSN |
Data mining,Static data,Computer science,Convex hull,Dynamic data,Cluster analysis,Generality | Conference | 2376-4066 |
Citations | PageRank | References |
0 | 0.34 | 0 |
Authors | ||
3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Liang Zhao | 1 | 39 | 5.13 |
Zhikui Chen | 2 | 692 | 66.76 |
Yi Yang | 3 | 0 | 0.34 |