PROBABILISTIC PROPERTIES OF THE INITIAL VALUES OF WEIGHTING FACTORS IN SYNCHRONIZED ARTIFICIAL NEURAL
Abstract
One of the most efficient ways for identical binary se quences generation is using methods of neural cryptography. The initial weight vestors values influence on speed of synchronization is analized. Equal probability of initial weight vestors motion directions is great advantage. On this base authors suppose new line of research conserned with improvement of network architecture and correction algorithm.
About the Authors
V. F. Golikov
Belarusian National Technical University
Belarus
Belarus
N. V. Brich
Belarusian National Technical University
Belarus
Belarus
References
1. Kanter, I. The Theory of Neural Networks and Cryptography, Quantum Computers and Computing / I. Kanter, W.Kinzel.−2005. – Vol. 5, n.1. – P. 130−140.
2. Kinzel, W. Neural Cryptography / W.Kinzel, / I. Kanter // 9th International Conference on Neural Information Processing, Singapore, 2002.
Review
For citations:
Golikov V.F., Brich N.V. PROBABILISTIC PROPERTIES OF THE INITIAL VALUES OF WEIGHTING FACTORS IN SYNCHRONIZED ARTIFICIAL NEURAL. «System analysis and applied information science». 2013;(1-2):33-37. (In Russ.)
Views:
2387
Email this article 