Abstract. 

Fuzzy systems with fuzzy inputs can be used in tasks where it is necessary to make predictions for data objects that have qualitative characteristics. However, building an optimal block of rules for such a system may be non-trivial, including due to the requirement to have a certain depth of knowledge in the subject area. In this situation, there is a need to automate the process of compiling the rule base, that is, to build a machine learning algorithm. In this paper, we propose to use a genetic (evolutionary) algorithm as such an algorithm. It describes both the specifics of using this family of algorithms for training a fuzzy system, and the features of parallel implementation of the learning process using CUDA technology.

Keywords: 

a logical type of inference, a decomposition theorem, evolutionary learning, parallel computations.

PP. 60-69.

DOI 10.14357/20718632210306

 

References

1. Rutkowski L. 2009. Methods and techniques of artificial intelligence. PWN. 452 p.

2. Zadeh L. 1976. Ponyatie lingvisticheskoy peremennoy i ego primenenie k prinyatiyu pribligennih resheniy [The concept of a linguistic variable and its application to approximate decision making]. Mir. 168 p.

3. A. N. Borisov, A. V. Alekseev, O. A. Krumberg. 1982. Decision models based on a linguistic variable. Riga: Zinatne. 256 p.

4. Wen-Ruey Hwang, W. E. Thompson. 1994. Design of intelligent fuzzy logic controllers using genetic algorithm.

5. C. L. Karr and E. J. Gentry. 1993. Fuzzy control of ph using genetic algorithms. IEEE Transactions on Fuzzy Systems   1(1):46–52 doi: 10.1109/TFUZZ. 1993.390283

6. Michael A. Lee and Hideyuki Takagi. 1993. Dynamic control of genetic algorithms using fuzzy logic techniques.In Proceedings of the 5th International Conference on Genetic Algorithms, page 76–83, San Francisco, CA, USA. Morgan Kaufmann Publishers Inc. ISBN 1558602992. doi: 10.5555/645513.657425.

7. D. Dubois and A. Prad. 1990. Theory of possibilities. Applications to the representation of knowledge in computer science. Radio and communications, Moscow. ISBN 5- 256-00184-1. doi: 10.1007/3-540-45493-4 24.

8. V. G. Sinuk, M. V. Panchenko. 2017. Methody nechetkogo vivoda dlya odnogo klassa system MISOstruktury pri nechetkih vhodah [Fuzzy inference method for a one class of MISO structure systems with fuzzy inputs.]. Iskustvenniy intellect i prinyatie resheniy [Artificial intelligence and decision making]. 4:33-39.

9. V. G. Sinuk, E. V. Pivnenko. 2006. Ob analiticheskom vichislenii nechetkogo znacheniya istinnosti [On analytical calculation of a fuzzy truth value]. Sbornik trudov Vserossiyskoy nauchnoy konferencii po nechetkim sistemam i myagkim vichisleniyam (NSMV-2006)[Proceedings of the all-Russian scientific conference on fuzzy systems and soft computing (NSMV-2006)]. 129-133.

10. D. A. Kucenko, V. G. Sinuk. 2008. Algoritmi nahogdeniya CP pri kusochno-lineynom predstavlenii funkciy prinadlegnosti[Algorithms for finding CP in piecewise linear representation of membership functions]. Sbornik trudov vtoroy Vserossiyskoy nauchnoy konferencii po nechetkim sistemam I myagkim vichisleniyam (NSMV- 2008)[Proceedings of the second all-Russian scientific conference on fuzzy systems and soft computing (NSMV- 2008)]. 87-92.

11. Yuhui Shi, R. Eberhart, and Yaobin Chen. 1999. Implementation of evolutionary fuzzy systems. IEEE Transactions on Fuzzy Systems, 7(2):109–119. doi: 10.1109/91.755393.

12. Jason Sanders, Edward Kandrot. 2010. CUDA by example: An introduction to general-purpose GPU programming. Addison-Wesley Professional. 320 p.

13. CUDA Programming Guide. Available at:

https://docs.nvidia.com/cuda/cuda-c-programmingguide/

index.html (accessed March 3, 2021)

14. CUDA Best Practices Guide. Available at:

https://docs.nvidia.com/cuda/cuda-c-best-practicesguide/

index.html (accessed March 3, 2021)

15. Balance scale data set. Available at:

https://archive.ics.uci.edu/ml/datasets/Balance+Scale (accessed

March 3, 2021)