The Neo-Fuzzy Neural Network Structure Optimization Using The GMDH For The Solving Forecasting And Classification Problems. Yevgen Viktorov, Maria Samarina, Elena Pavlikovskaya, Zaychenko Yurii, Yevgeniy Bodyanskiy

Abstract. The problem of Neo-Fuzzy Neural Network structure optimization is considered in this paper. For its solution Group Method of Data Handling (GMDH) is suggested and the algorithm of structure optimization is described.. The experimental investigations were carried out and their results accuracy of forecasting by optimally constructed Neo- Fuzzy Neural Network and network with multilayer feedforward architecture are presented and compared. Also a classification problem was solved and results are given in this paper to show that proposed self-organized architecture is capable to perform classification as well as forecasting.

Keywords. Artificial neural networks, neo-fuzzy neuron, group method of data handling, structure optimization, stock prices forecasting.

References.

1. Yamakawa T., Uchino E., Miki T., Kusanagi H. A neo fuzzy neuron and its applications to system identification and prediction of the system behavior // Proc. 2-nd Int.Conf. on Fuzzy Logic and Neural Networks “LIZUKA– 92”. Lizuka, Japan. – 1992. – P. 477–483.

2. Uchino E., Yamakawa T. Soft computing based signal prediction, restoration and filtering // Intelligent Hybrid Systems: Fuzzy Logic, Neural Networks and Genetic Algorithms. Ed. Da Ruan. – Boston: Kluwer Academic Publisher. – 1997. – P. 331–349.

3. Miki T., Yamakawa T. Analog implementation of neo-fuzzy neuron and its on-board learning // Computational Intelligence and Applications. Ed. N. E. Mastorakis. – Piraeus: WSES Press. – 1999. – P. 144–149.

4. Bodyanskiy Ye., Kokshenev I., Kolodyazhniy V. An adaptive learning algorithm for a neo-fuzzy neuron // Proc. 3rd Int. Conf. of European Union Soc. for Fuzzy Logic and Technology (EUSFLAT 2003). – Zittau, Germany. – 2003. – P. 375–379.

5. Kolodyazhniy V., Bodyanskiy Ye., Otto P. Universal approximator employing neo-fuzzy neurons // In “Computational Intelligence: Theory and Applications.” Ed. by B. Reusch. – Berlin-Heidelberg: Springer. – 2005. – P. 631–640.

6. Kaczmarz S. Angenaeherte Ausloesung von Systemen Linearer Gleichungen // Bull. Int. Acad. Polon. Sci. – 1937. – Let. A. – S. 355–357.

7. Kaczmarz S. Approximate solution of systems of linear equations // Int. J. Control. – 1993. – 53. – P. 1269–1271.

8. Widrow B., Hoff Jr. M. E. Adaptive switching circuits // 1960 URE WESCON Convention Record. – N.Y.: IRE. – 1960. – Part 4. – P. 96–104.

9. Bodyanskiy Ye., Kolodyazhniy V. Adaptive nonlinear control using neo-fuzzy model / Eds. by O. Sawodny, P. Scharff “Synergies between Information Processing and Automation”. – Aachen: Shaker Verlag, 2004. – P. 122- 127.

10. Zaychenko Yu. “The Fuzzy Group Method of Data Handling and Its Application for Economical Processes Forecasting” - Scientific Inquiry, - Vol. 7, No.1, June, 2006 - p.83-96.

11. Zaychenko Yu. Fuzzy method of inductive modeling in problems of macroeconomic indexes forecasting. System researches and informational technologies, #3 of 2003, p. 25-45.

12. Zaychenko Yu. P., and Zayetz I.O. The synthesis and adaptation of fuzzy forecasting model based of self- organization method. Science News of NTUU “KPI”, #2 of 2001.

13. Zaychenko Yu. P., and Zayetz I.O. Research of different types of partial descriptions in problems of synthesis of fuzzy forecasting models, Science works of Donetsk NTU, vol. 47, p. 341-349.

14. Zaychenko Yu. P. Comparative analysis of forecasting models built using distinct and fuzzy GMDH with different algorithms of fuzzy forecasting models generation. Materials of international seminar of inductive modeling IWIM 2005.