Abstract
In this paper we study some dynamical aspects of the synchronous update of Neural Networks. We exhibit Lyapunov Functional for discrete neural networks (binary states) and also for continuous local rules. By doing so we prove, in the context of binary networks, the convergence to cycles of period one or two for symmetric weights. Also we prove that “almost symmetric” Neural Networks admit large cycles (non-bounded in the size, n, of the array) and that the synchronous update with symmetric weights may have an exponential transient time in order to reach a fixed point. Finally, for continuous local rules we give the explicit expression of Lyapunov operators for synchronous and sequential iterations and we characterize the network dynamics for periodic trajectories.
Partially supported by FNC/89, DTI U. Chile, EHEI U. Paris-V and F. Andes-Chile/89
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Bienenstock E., Fogelman-Soulie F., Weisbuch G., Disordered Systems and Biological Organisation, Procc Les Houches, NATO ASI series, Springer Verlag, 1986
Fogelman-Soulie F., Robert Y., Tchuente M., Automata Networks in Computer Science, Manchester Univ. Press, 1987
Fogelman-Soulie F., Goles E., Weisbuch G., Transient length in sequential iteration of threshold functions, Disc. Appl. Maths., 6, 95–98, 1983
Fogelman-Soulie F., Goles E., Martinez S., Mejia C., Energy functionals in Neural Networks with continuous local functions, Res. Rep. EHEI, Paris-V, 1988, submitted to Complex Systems
Goles E., Martinez S., Dynamics on Generalized Neural Networks, to appear in Kluber Pub. (ex Reidel), 1990
Goles E., Olivos J., Comportement périodique des fonctions a seuil binaires et applications Disc. Appl. Maths., 3, 93–105, 1981
Goles E., Fixed point behaviour of threshold functions, SLAM J. on Disc, and Alg. Meths., 4, 529–531, 1982
Goles E., Dynamics on Positive Automata Networks, Theor. Comp. Sci., 41, 19–32, 1985
Goles E., Antisymmetric Neural Networks, Disc. Applied Maths., 13, 97–100, 1986
Goles E., Martinez S., Properties on positive functions and the dynamics of associated automata networks, Disc. Appl. Maths., vol 18, 39–46, 1987
Goles E., Martinez S., Lyapunov functionals for Automata Networks defined by ciclically monotone functions, Res. Rep., Dep. Mat., U. de Chile, 1987, submitted to SIAM J. Disc. Maths.
Goles E., Vichniac G., Lyapunov functions for parallel neural networks, in Neural Networks for Computing, Snowbird 1986, Denker ed., Am. Inst. Phys., 151, 165–181, 1986
Goles E., Fogelman-Soulie F., Pellegrin D., Decreasing energy functions as a tool for studying threshold functions, Disc. Applied Maths., 12, 261–277, 1985
Goles E., Olivos J., The convergence of symmetric threshold automata, Inf. and Control, vol 51, 2, 98–100, 1981
Goles E., Martinez S., Exponential transient classes of symmetric neural networks for synchronous and sequential updating, Res Rep. Dep. Matematicas, U. de Chile, 1989, submitted to Neural Networks.
Hopfield J.J., Neural Networks and physical systems with emergent collective computational abilities, Proc. Nat. Acad. Sci, USA 79, 2554–58, 1982
Hopfield J.J., Tank D.W., Neural Computation of Decisions in Optimization Problems, Biol. Cybern., 52, 141–152, 1985
Klenne S.C., Representation of events in nerve nets and finite automata, in Automata Studies, Shannon C. E., Mac Carthy J. eds., Princeton Univ. Press, 1956
McCulloch W., Pitts W., A logical calculus of the ideas immanent in nervous activity, Bull, of Math. Biophys., 5, 1943
Rosenblatt F., Principles of Neurodynamics, Spartan, Washington, 1962
Rumelhart D. E., Hinton G., Williams R. J., Learning Integral Representation by Error Propagation, in Parallel Distributed Proccesing, MIT Press, Vol 1, 318–362, 1986
Shingai R., Maximun period of 2-dimensional neural networks, Inf. and Control 41, 324–341, 1979
Tchuente M., Sequential simulation of parallel iterations and applications, Theor. Comp. Sci., 48, 135–144, 1986
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 1990 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Goles, E. (1990). Neural Networks Dynamics. In: Soulié, F.F., Hérault, J. (eds) Neurocomputing. NATO ASI Series, vol 68. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-76153-9_11
Download citation
DOI: https://doi.org/10.1007/978-3-642-76153-9_11
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-642-76155-3
Online ISBN: 978-3-642-76153-9
eBook Packages: Springer Book Archive