Learning Algorithms for Fuzzy Inference Systems Composed of Double- and Single-Input Rule Modules

Most of self-tuning fuzzy systems, which are
automatically constructed from learning data, are based on the
steepest descent method (SDM). However, this approach often
requires a large convergence time and gets stuck into a shallow
local minimum. One of its solutions is to use fuzzy rule modules
with a small number of inputs such as DIRMs (Double-Input Rule
Modules) and SIRMs (Single-Input Rule Modules). In this paper,
we consider a (generalized) DIRMs model composed of double
and single-input rule modules. Further, in order to reduce the
redundant modules for the (generalized) DIRMs model, pruning and
generative learning algorithms for the model are suggested. In order
to show the effectiveness of them, numerical simulations for function
approximation, Box-Jenkins and obstacle avoidance problems are
performed.

Authors:

• Hirofumi Miyajima
• Kazuya Kishida
• Noritaka Shigei
• Hiromi Miyajima

Keywords:

• Box-Jenkins’s problem
• Double-input rule module
• Fuzzy inference model
• Obstacle avoidance
• Single-input rule module.

References:

[1] B. Kosko, Neural Networks and Fuzzy Systems: A Dynamical Systems
Approach to Machine Intelligence, Prentice Hall, Englewood Cliffs, NJ,
1992.
[2] C. Lin, and C. Lee, Neural Fuzzy Systems, Prentice Hall, PTR, 1996.
[3] M.M. Gupta, L. Jin, and N. Homma, Static and Dynamic Neural
Networks, IEEE Press, 2003.
[4] J. Casillas, O. Cordon, F. Herrera, and L. Magdalena, “Accuracy
Improvements in Linguistic Fuzzy Modeling,” Studies in Fuzziness and
Soft Computing, Vol.129, Springer, 2003.
[5] B. Liu, Theory and Practice of Uncertain Programming, Studies in
Fuzziness and Soft Computing, Vol.239, Springer, 2009.
[6] O. Cordon, “A historical review of evolutionary learning methods for
Mamdani-type fuzzy rule-based systems: Designing interpretable genetic
fuzzy systems,” Journal of Approximate Reasoning, 52, pp.894-913,
2011.
[7] K. Kishida, H. Miyajima, M. Maeda, and S. Murashima, “A Self-tuning
Method of Fuzzy Modeling using Vector Quantization,” Proceedings of
FUZZ-IEEE’97, pp.397-402, 1997.
[8] W. Pedrycz, and H. Izakian, “Cluster-Centric Fuzzy Modeling,” IEEE
Trans. on Fuzzy Systems, Vol.22, Issue 6, pp. 1585-1597, 2014.
[9] S. Fukumoto, and H. Miyajima, “Learning Algorithms with
Regularization Criteria for Fuzzy Reasoning Model,” Journal of
Innovative Computing, Information and Control, vol.1, no.1, pp.249-263,
2006.
[10] N. Yubazaki, J. Yi, and K. Hirota, “SIRMS (Single Input Rule Modules)
Connected Fuzzy Inference Model,” J. Advanced Computational
Intelligence, 1, 1, pp.23-30, 1997.
[11] N. Shigei, H. Miyajima, and S. Nagamine, “A Proposal of Fuzzy
Inference Model Composed of Small-Number-of-Input Rule Modules,”
Proc. of Int. Symp. on Neural Networks: Advances in Neural Networks
- Part II, pp.118-126, 2009.
[12] S. Miike, H. Miyajima, N. Shigei, and K. Noo, “Fuzzy Reasoning
Model with Deletion of Rules Consisting of Small-Number-of-Input Rule
Modules,” Journal of Japan Society for Fuzzy Theory and Intelligent
Informatics, pp.621-629, 2010 (in Japanese).
[13] H. Miyajima, N. Shigei, and H. Miyajima, “An Application of
Fuzzy Inference System Composed of Double-Input Rule Modules to
Control Problems,” Proceedings of the International MultiConference of
Engineers and Computer Scientists 2014, Vol I, pp.23-28, 2014.
[14] H. Miyajima, N. Shigei, and H. Miyajima, “Some Properties on Fuzzy
Inference Systems Composed of Small Number of Input Rule Modules,”
Advances in Fuzzy Sets and Systems, Vol.20, pp.155-175, 2015.
[15] G.E.P. Box, G.M. Jenkins, Time series analysis, forecasting and
control,second ed., Holden Day, San Francisco, CA, 1970.
[16] E. Kim, M. Park, S. Ji, and M. Park, “A new approach to fuzzy
modeling,” IEEE Trans. Fuzzy Systems, vol. 5, no. 3, pp. 328-337, 1997.
[17] C. Li, J. Zhou, B. Fu, P. Kou, and J. Xiao, “T-S fuzzy model
identification with a gravitational search-based hyperplane clustering
algorithm,” IEEE Trans. Fuzzy Syst., vol.20, no.2, pp. 305-317, 2012.