Abstract
This chapter brings together mathematical and physical tools which lie at the root of a derivation of the learning rules for dynamic recurrent neural networks. It starts with the basic problem in the calculus of variations of how to determine a set of functions y i /(x), i = 1,2,…, n, which minimize (or maximize) the integral of some functional of these functions. This problem is generalized by the introduction of constraints such that the solutions also have to satisfy the constraint equations. This results in the method of the ‘Lagrange undetermined multiplier functions’. Application of these methods in physics will be discussed after the introduction of Hamilton’s Principle. The Lagrangian and Hamiltonian energy functions together with the canonical equations of motion of Hamilton are introduced. The way in which Pontryagin applied these methods in the construction of his ‘Minimum Principle of Pontryagin’ for the solution of general optimal control problems is presented. The chapter ends with the derivation of the learning rules.
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
Boltyanskii, V.G., Gamkrelidze, R.V., Mishchenko, E.F. and Pontryagin, L. S. (1966): The maximum principle in the theory of optimal processes of control. In: Optimal and self-organizing control, R. Oldenburger, Ed. Cambridge, MA: MIT Press
Draye, J-Ph.S., Pavisic, D.A., Cheron, G.A. and Libert, G.A. (1995): Adaptive time constants improve the prediction capability of recurrent neural networks. Neural Processing Letters 2(3), 12–16
Draye, J-Ph.S., Pavisic, D.A., Cheron, G.A. and Libert, G.A. (1996): Dynamic recurrent neural networks: A dynamical analysis. IEEE Transactions on Systems, Man and Cybernetics — Part B: Cybernetics 26, 692–706
Fan, L.T. (1966): The continuous maximum principle. A study of complex systems optimization. J Wiley and Sons
Gamkrelidze, R.V. (1978): Principles of optimal control theory. Plenum Press.
Kibble, T.W.B. and Berkshire, F.H. (1996): Classical mechanics. Addison Wesley Longman Ltd
Landau, L.D. and Lifschitz, E.M. (1991): Mechanics. (Vol. 1 of the Course of Theoretical Physics). Pergamon Press
Leech, J.W. (1958): Classical mechanics. Methuen’s monographs on physical subjects. Methuen and Co Ltd
Pearlmutter, B.A. (1989): Learning state space trajectories in recurrent neural networks. Neural Computation 1, 263–269
Pearlmutter, B.A. (1995): Gradient calculations for dynamic recurrent neural networks: A survey. IEEE Transactions on Neural Networks 6, 1212–1228
Pontryagin, L.S., Boltyanskii, V.G., Gamkrelidze, R.V. and Mishchenko, E. F. (1962): The mathematical theory of optimal processes. J Wiley and Sons
Schultz, D.G. and Melsa, J.L. (1967): State functions and linear control systems. McGraw-Hill
Thornton, M. (1995): Classical dynamics of particles and systems. Saunders College Publishing. Hartcourt Brace College Publishers
Werbos, P.J. (1988): Generalization of backpropagation with application to a recurrent gas market model. Neural Networks 1, 339–356
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2001 Springer Science+Business Media Dordrecht
About this chapter
Cite this chapter
Mastebroek, H.A.K. (2001). A Derivation of Learning Rules for Dynamic Recurrent Neural Networks. In: Mastebroek, H.A.K., Vos, J.E. (eds) Plausible Neural Networks for Biological Modelling. Mathematical Modelling: Theory and Applications, vol 13. Springer, Dordrecht. https://doi.org/10.1007/978-94-010-0674-3_4
Download citation
DOI: https://doi.org/10.1007/978-94-010-0674-3_4
Publisher Name: Springer, Dordrecht
Print ISBN: 978-94-010-3864-5
Online ISBN: 978-94-010-0674-3
eBook Packages: Springer Book Archive