Abstract
A digital neuron circuit suggested in this work consists only of simple logic gates, several adders, and several random number generators. This makes it possible to reduce hardware costs and increase the number of synapses. A mathematical model and experiments made with a model program are described. The Rprop optimization algorithm is used for learning.
Similar content being viewed by others
REFERENCES
Tomlinson, M.S. and Walker, D.J., DNNA: A Digital Neural Networks Architecture, Proc. Int. Neural Networks Conf. (INNC-90), 1990, vol. 2, pp. 589-592.
Mamatov, Yu.A., Bulychev, S.F., Karlin, A.K., et al., Digital Implementation of a Stream Neuron, Radiotekh. Elektron. (Moscow), 1995, no. 11, pp. 1652-1660.
Mamatov, Yu.A., Bulychev, S.F., Karlin, A.K., et al., Circuit Representation of Digital Pulse-Stream Neuron Models, Mikroelektronika, 1996, vol. 25, no. 1, pp. 3-8.
Timofeev, E.A., Simulation of a Neuron Representing Information as a Stream Flux Density, Avtom. Telemekh., 1997, no. 3, pp. 190-199.
Riedmiller, M. and Braun, H., A Direct Adaptive Method for Faster Backpropagation Learning: The RPROP Algorithm, Proc. IEEE Int. Conf. on Neural Networks (ICNN), San Francisco, 1993, pp. 586-591.
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Luk'yanov, A.V. A Circuit-Design Model of a Digital Neuron Operating with the Mean Value of a Stochastic Stream. Russian Microelectronics 30, 54–58 (2001). https://doi.org/10.1023/A:1009425926051
Issue Date:
DOI: https://doi.org/10.1023/A:1009425926051