An extended neuron model for efficient timeseries generation and prediction
Modelling of spatio-temporal patterns with neural networks is an important task for a large number of applications which require adaptive control. In this paper, the use of an extended neuron model in neural networks is proposed to achieve a given dynamic network behaviour. The new neuron model is based on the implementation of exponential excitation decay and the introduction of temporal refractoriness of the neuron output as observed in biological nerve cells. A learning algorithm based on error-backpropagation for the resulting network is derived. A benchmark test on prediction of the chaotic Macky-Glass differential equation and real-live experiments with controlling the movement of a walking machine leg are performed. The results suggest superior time-series modelling ability of the presented approach in terms of network trainability and computation complexity.
Unable to display preview. Download preview PDF.
- 1.Kinouchi M, Hagiwara, M: “Learning Temporal Sequences by Complex Neurons with Local Feedback”, Proc. 1995 IEEE Int. Conf. Neural Networks, 1995, pp. 3165–3169Google Scholar
- 2.Jordan M. I.: “Attractor Dynamics and Parallelism in a Connectionist Sequential Machine”, Proc. 8th Annual Conf. of the Cognitive Science Society, 1986Google Scholar
- 3.Day S. P., Davenport M. R.: “Continuous-Time Temporal Back-Propagation with Adaptable Time Delays”, IEEE Transact. Neural Networks, 1993, Vol. 4 No. 2, pp. 348–354Google Scholar
- 4.Aihara K., Takabe T., Toyoda M.: “Chaotic Neural Networks”, Physics Letters, 1990, Vol. 144 No. 6/7Google Scholar
- 5.Watanabe M., Aihara K., Kondo S.: “Spatio-temporal Summation and Selforganization in Chaotic Neural Networks”, Proc. 1995 IEEE Int. Conf. Neural Networks, 1995, pp 3150–3153Google Scholar
- 6.Nossek J. A., Nachbar P, Schuler A. J.: “Comparison of Learning Algorithms for Feedforward Neural Nets”, Proc. 1996 IEEE Int. Conf. Circuits and Systems, 1996, pp. 380-384Google Scholar