International Conference on Information Technology and Computer Science, 3rd (ITCS 2011)
112 One of the Adaptive Clustering Algorithm Based on Typical and Stochastic Data Sets
Download citation file:
- Ris (Zotero)
- Reference Manager
Aiming at resolving the deficits of traditional clustering algorithm, the paper puts forward an adaptive clustering algorithm. Based on k-means algorithm, this algorithm combines the idea of automatically gathered of ant colony algorithm and the idea of probability search of simulated annealing algorithm. Apart from marketing research, clustering algorithm is also widely utilized in pattern recognition, data analysis and other important domains. Through the test of some examples, we evaluated the adaptive algorithm on the typical data sets and the stochastic data sets. In comparison with other three algorithms, adaptive algorithm can not only automatically calculate the number k of clusters, but also cut down on the repetitions to enhance the performance. Meanwhile, the firmness of algorithm is optimal.