A clustering algorithm based on swarm intelligence
Authors:
- Wu binWBWu bin
- Shi ZhongzhiSZShi Zhongzhi
- 13 November 2002
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 3, 58-66
Abstract
This paper focuses on swarm intelligence based clustering algorithm. A clustering algorithm based on swarm intelligence is systematically proposed. It derived from a basic model interpreting ant colony organization of cemeteries. Some important concepts, such as swarm similarity, swarm similarity coefficient and probability conversion function are also proposed. A simplified probability conversion function is given for simplifying adaptation of parameters, meanwhile the importance of swarm similarity coefficient for the algorithm is analyzed. Experimental results show the good performance of the clustering algorithm.Keywords
This publication has 6 references indexed in Scilit:
- – Ant SystemFuture Generation Computer Systems, 2000
- Swarm Intelligence1999
- A Stochastic Heuristic for Visualising Graph Clusters in a Bi-Dimensional Space Prior to PartitioningJournal of Heuristics, 1999
- AntNet: Distributed Stigmergetic Control for Communications NetworksJournal of Artificial Intelligence Research, 1998
- Diversity and Adaptation in Populations of Clustering Ants1994
- The Dynamics of Collective Sorting: Robot-Like Ants and Ant-Like Robots1991