A.V. Novikov - Post-graduate Student, Department of Computer Systems and Software Technologies, St.-Petersburg State Polytechnical University. E-mail: spb.andr@yandex.ru E.N. Benderskaya - Ph.D.(Eng.), Associate Professor, Department of Computer Systems and Software Technologies, St.-Petersburg State Polytechnical University. E-mail: helen.bend@gmail.com

The oscillatory network Sync that uses the phase oscillator Kuramoto is able to solve problems of cluster analysis using processes of synchronization when one synchronous ensemble of oscillator corresponds to only one cluster. The convergence rate of process of synchronization depends on number of oscillators and radius connectivity. This paper presents study results of synchronization processes in oscillatory networks that based on the Kuramoto model and proposed double layer oscillatory network for cluster analysis that resolves described two problems. The oscillatory network SYNC-SOM provides accurate solution of the clustering problems than traditional algorithms. The network is relevant for problems where accurate and fast solution is required. Also presented experimental results of comparison between various algorithms and advantages of the proposed algorithm.

 

1. Anil K., Dubes J.C., Dubes R.C. Algorithm for Clustering Data / Prentice Hall, Englewood Cliffs, New Jersey. 1998. P. 304.
2. Basar E. Brain function and oscillations / Springer-Verlag New York. 1998. P. 363.
3. Bohm C., Plant C., Shao J., Yang Q. Clustering by synchronization // KDD - 10 Proceedings of the 16th ACM SIGKDD international conference of Knowledge discovery and data mining, 2010. P. 583-592.
4. Ester M., Kriegel H., Sander J., Xu. X. A Density Based Algorithm for Discovering Clusters in Large Spatial Data Sets with Noise // Proceedings of the 2nd International Conference on Knowledge Discovery and Data Mining. 1996. P. 226-231.
5. Guha S., Rastogi R., Shim K. ROCK: A Robust Clustering Algorithm for Categorical Attributes // Proceedings of the 15th International Conference on Data Engineering. 1999. P. 512-521.
6. Haken E. Brain Dynamics / Springer-Verlag Berlin Neidelberg. 2007. P. 238.
7. Haykin S. Neural networks: A comprehensive foundation / Macmillan College Publishing Company New York, 2nd edition. 1999. P. 842.
8. Kohonen T. Self-Organizing Maps / Springer-Verlag Berlin, 3rd edition. 2001. P. 501.
9. Kuramoto Y. Chemical Oscillations Waves, and Turbulence / Springer-Verlag Berlin Neidelberg New York Tokyo. 1984. P. 157.
10. MacQueen J.B. Some Methods for Classification and Analysis of Multivariate Observations // Proceedings of the 5th Berkley Symposium Math. Statistics and Probability. 1967. V. 1. P. 281-297.
11. Miyano T., Tsutsui T. Data Synchronization as a Method of Data Mining // International Symposium on Nonlinear Theory and its Applications. 2007.
12. Novikov A.V., Benderskaya E.N.The Oscillatory Neural Networks Based on Kuramoto Model for Cluster Analysis // Proceedings of the 11th International Conference \"Pattern Recognition and Image Analysis: New Information Technologies\". 2013. V. 1. P. 106-109.
13. Novikov A.V., Benderskaya E.N.SYNC-SOM Double-layer Oscillatory Network for Cluster Analysis // 3rd International Conference on Pattern Recognition Applications and Methods, Proceedings, ESEO, Angers, Loire Valley. France. 6-8 March. 2014. P. 305-309.
14. Sudipto G., Rajeev R., Kyuseok S. CURE: An Efficient Clustering Algorithm for Large Databases // Proceedings of the SIGMOD\'98. 1998. P. 73-84.
15. Ultsch A. Clustering with SOM: U*C // Workshop on Self Organizing Feature Maps. 2005. P. 31-37.
16. Novikov A.V., Benderskaja E.N. Reshenie zadach klasternogo analiza na osnove oscilljatornykh nejjronnykh setejj // Nejjrokompjutery: razrabotka, primenenie. 2013. № 12. S. 31-36.
17. Novikov A.V., Benderskaja E.N. Nejjrosetevye metody reshenija zadach klasternogo analiza // Nejjrokompjutery: razrabotka, primenenie. 2014. № 2. S. 48-53.