Node priority guided clustering algorithm
- Publication Type:
- Journal Article
- Kongzhi yu Juece/Control and Decision, 2011, 26 (6)
- Issue Date:
Density-based clustering algorithms have the advantages of clustering with arbitrary shapes and handling noise data, but cannot deal with unsymmetrical density distribution and high dimensionality dataset. Therefore, a node priority guided clustering algorithm(NPGC) is proposed. A direct K neighbor graph of dataset is set up based on KNN neighbor method. Then the local information of each node in graph is captured by using KNN kernel density estimate method, and the node priority is calculated by passing the local information through graph. Finally, a depth-first search on graph is applied to find out the clustering results based on the local kernel degree. Experiment results show that NPGC has the ability to deal with unsymmetrical density distribution and high dimensionality dataset.
Please use this identifier to cite or link to this item: