Abstract
Neural networks are systems of interconnected processors mimicking some of the brain functions. After a rapid overview of neural computing, the thermodynamic formalism of the learning procedure is introduced. Besides its use in introducing efficient stochastic learning algorithms, it gives an insight in terms of information theory. Main emphasis is given in the information restitution process; stochastic evolution is used as the starting point for introducing statistical mechanics of associative memory. Instead of formulating problems in their most general setting, it is preferred stating precise results on specific models. In this report are mainly presented those features that are relevant when the neural net becomes very large. A survey of the most recent results is given and the main open problems are pointed out.
Work partially supported by EU network CHRX-CT93-0411.
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
Abbott, L.F., Learning in Neural Network Memories, Network 1, 105–122 (1990).
Aizenman, M., J.L. Lebowitz, D. Ruelle, Some Rigorous Results on the Sherrington-Kirkpatrick Model, Commun. Math. Phys., 112, 3–20 (1987).
Albeverio, S., B. Tirozzi, B. Zegarlinski, Rigorous Results for the Free Energy in the Hopfield Model, Commun. Math. Phys., 150, 337–373 (1992).
Amaro de Matos, J.M.G., A.E. Patrick, V.A. Zagrebnov, Random Infinite Volume Gibbs States for the Curie-Weiss Random Field Ising Model, J. Stat. Phys., 66, 139–164 (1992).
Amit, D.J., Modelling Brain Function, Cambridge University Press, Cambridge (1989).
Amit, D.J., G. Gutfreund, H. Sompolinsky, Statistical Mechanisms of Neural Networks near Saturation, Ann. Phys., 173, 30–67 (1987).
Anily, S., A. Federgruen, Simulated Annealing Methods with General Acceptance Probabilities, J. Appl. Prob. 24, 657–667 (1968).
Aubin, J.P., Mathematical Methods for Neural Networks, lecture notes of a COMETT graduate module held in Les Houches, 16–29 March (1992).
Azencott, R., Simulated Annealing, Séminaire Bourbaki No. 697, 1–15 (1988).
Ben Arous, G., A. Guionnet, Large Deviations for Langevin Spin Glass Dynamics, preprint DMI, École Normale Supérieure (1995).
Bernier, O., Stochastic Analysis of the Dynamics of a General Class of Synchronous Neural Networks, J. Phys. A: Math. Gen., 26, 6879–6892 (1993).
Billingsley, P., Convergence of Probability Measures, Wiley, New York (1968).
Bovier, A., Self-Averaging in a Class of Generalised Hopfield Models, J. Phys. A: Math. Gen. 27, 7069–7077 (1994).
Bovier, A., V. Gayrard, Rigorous Bounds on the Storage Capacity of the Dilute Hopfield Model, J. Stat. Phys. 69, 597–627 (1992).
Bovier, A., V. Gayrard, Rigorous Results on the Thermodynamics of the Dilute Hopfield Model, J. Stat. Phys., 72, 79–112 (1993).
Bovier, A., V. Gayrard, An almost sure Large Deviation Principle for the Hopfield Model, Weierstraß IAAS preprint (1995).
Bovier, A., V. Gayrard, P. Picco, Gibbs States of the Hopfield Model in the Regime of Perfect Memory, Prob. Th. Rel. Fields, 100, 329–363 (1994).
Bovier, A., V. Gayrard, P. Picco, Gibbs States of the Hopfield Model with Extensively Many Patterns, J. Stat. Phys., 79, 395–414 (1995).
Bovier, A., V. Gayrard, P. Picco, Large Deviation Principles for the Hopfield and the Kac-Hopfield model, Weierstraß IAAS preprint (1994).
Broeck, C. van den, Statistical Physics of Learning from Examples: a Brief Introduction, Acta Phys. Polon. B, 25, 903–923 (1994).
Cassandro, M., A. Galves, E. Olivieri, M. E. Vares, Metastable Behaviour of Stochastic Dynamics: a Pathwise Approach, J. Stat. Phys., 35, 603— (1984).
Changeux, J.-P., L’Homme Neuronal, Fayard, Paris (1983).
Collet, P., J.-P. Eckmann, Iterated Maps on the Interval as Dynamical Systems, Birkhäuser, Basel (1980).
Cottrell, M., Mathematical Analysis of a Neural Network with Inhibitory Coupling, Stoch. Proc. Appl., 40, 103–126 (1992).
Derrida, B., R.B. Griffiths, A Prügel-Bennett, Finite-Size Effects and Bounds for Perceptron Models, J. Phys. A: Math. Gen., 24, 4907–4940 (1991).
Dobrushin, R.L., The Description of a Random Field by Means of Conditional Probabilities and Condition of its Regularities, Th. Prob. Appl., 13, 458–486 (1968).
Fassnacht, C., A. Zippelius, A recognition and categorisation in a structured neural network with attractor dynamics, Network, 2, 63–84 (1992).
Ferrari, P.A., S. Martinez, P. Picco, A Lower Bound for the Memory Capacity in the Potts-Hopfield Model, J. Stat. Phys., 66, 1643–1652 (1992).
Folk, R., A. Kartashov, P. Lisonék, P. Paule, Symmetries in Neural Networks: a Linear Group Action Approach, J. Phys. A: Math. Gen., 26, 3159–3164 (1993).
Freidlin, M.I., A.D. Wentzell, Random Perturbations of Dynamical Systems, Springer-Verlag, Berlin (1984).
Gardner, E., The Space of Interactions in Neural Network Models, J. Phys. A: Math. Gen., 21, 257–270 (1987).
Gardner, E., B. Derrida, Optimal Storage Properties of Neural Network Models, J. Phys. A: Math. Gen., 21, 271–284 (1988).
Gayrard, V., The Thermodynamic Limit of the Potts-Hopfield Model for Infinitely Many Patterns, J. Stat. Phys., 68, 977–1011 (1992).
Georgii, H.-O., Gibbs Measures and Phase Transitions, Walter de Gruyter, Berlin (1988).
Girko, V.L., Random Matrices, Vishcha Shkola, Izdat. Kiev Univ., Kiev (1975).
Girko, V L, Limit Theorems for Maximal and Minimal Eignevalues of Random Matrices, Th. Prob. Appli., 35, 680–695 (1988).
Girko, V.L., Theory of Random Determinants, Kluwer, Dodrecht (1990).
Goles, E., S. Martinez, Neural and automata networks, Kluwer Academic Publ., Dodrecht (1990).
Goles, E., S. Martinez, Statistical Physics, Automata Networks, and Dynamical Systems, Kluwer Academic Publ., Dodrecht (1992).
Griniasty, M., M.V. Tsodyks, D.J. Amit, Conversion of Temporal Correlations between Stimuli to Spatial Correlations between Attractors, preprint Università di Roma 1 (1992).
Haario, H., E. Saksman, Simulated Annealing Process in General State Space, Adv. Appl. Prob., 23, 866–893 (1991).
Hebb, D., The Organisation of Behaviour: a Neurolophysiological Theory, Wiley, New York (1949).
Hemmen, J.L. van, D. Grensing, A. Huber, R. Kühn, Nonlinear Neural Networks I: General Theory, J. Stat. Phys., 50, 231–257 (1988).
Hemmen, J.L. van, D. Grensing, A. Huber, R. Kühn, Nonlinear neural networks II: information processing, J. Stat. Phys., 50, 259–293 (1988).
Hertz, J., A. Krogh, R. Palmer, Introduction to the theory of neural computation, Addison-Wesley, Redwood City CA (1991).
Hopfield, J.J., Neural Networks and Physical Systems with Emergent Collective Computational Abilities, Proc. Natl. Acad. Sci., 79, 2554–2558 (1982).
Kallianpur, G., Weak Convergence of Stochastic Neuronal Models, in Stochastic Methods in Biology, Nagoya (1985), M. Kimura, G. Kallianpur, and T. Hida (eds.).
Khinchin, A.I., Mathematical Foundations of Information Theory, Dover, New York (1957).
Koch, H., A Free Energy Bound for the Hopfield Model, J. Phys. A: Math. Gen., 26, L353 — L355 (1993).
Koch, H., J. Piasko, Some Rigorous Results on the Hopfield Neural Network Model, J. Stat. Phys., 55, 903–928 (1989).
Kolmogorov, A.N., Foundations of Probability Theory, Chelsea Publishing, New York (1950).
Komlós, J., R. Paturi, Convergence Results in a Autoassociative Memory Model, Neural Networks, 1, 239–250 (1988).
Kosko, B., Neural Networks and Fuzzy Systems, Prentice-Hall, London (1992).
Krauth, W., M. Mézard, Storage Capacity of Memory Networks with Binary Couplings, J. Physique, 50, 3057–3066 (1989).
Kuramoto, Y., Chemical Oscillation, Waves, and Turbulence, Springer-Verlag, Berlin (1984).
Lanford III, O.E., D. Ruelle, Observables at Infinity and States with Short Range Correlations in Statistical Mechanics, Commun. Math. Phys. 13, 194215 (1969).
Lang, S., Real and Functional Analysis, Springer-Verlag, Berlin (1993).
Ledoux, M., M. Talagrand, Probability in Banach Spaces, Springer-Verlag, Berlin (1991).
Little, W., G. Shaw, Analytic Study of the Memory Storage Capacity of a Neural Network, Math. Biosc., 39, 281–290 (1978).
Loukianova, D., Capacité de Mémoire dans le Modèle de Hopfield, C. R. Acad. Sci. Paris, 318, 157–160 (1994).
Loukianova, D., Étude Rigoureuse du Modèle de Mémoire Associative, PhD Thesis, Université de Paris VII, presented on 2 December (1994).
MacEliece, R., E. Posner, E. Rodemich, S. Venkatesh, The Capacity of the Hopfield Associative Memory, IEEE Trans. Inf. Theory, 33, 461–482 (1987).
Maes, C., V. Zagrebnov, On the Parallel Dynamics of a Multi-Layered Perceptron, (1991).
Martínez, S., Introduction to Neural Networks: Storage Capacity and Optimisation, proceedings CIMPA school on “Dynamical and disordered systems”, Temuco (1992).
McCulloch, W.S., W. Pitts, A Logical Calculus of the Ideas Immanent in Nervous Activity, Bull. Math. Biophys. 5, 115–133 (1943).
Mézard, M., G. Parisi, M.A. Virasoro, Spin-glass Theory and Beyond, World scientific, Singapore (1988).
Minlos, R., Gibbs’ Limit Distribution, Funct. Anal. Appl., 2, 60–73; 3, 40–53 (1967).
Miyashita, Y., Neuronal Correlate of Visual Associative Long-Term Memory in the Primate Temporal Cortex, Nature, 335, 817–819 (1988).
Monasson, R., Properties of Neural Network Storing Spatially Correlated Patterns, J. Phys. A: Math. Gen., 25, 3701–3720 (1992).
Müller, B., J. Reinhardt, Neural Networks, Springer-Verlag, Berlin (1990).
Newman, C., Memory Capacity in Neural Networks, Neural Networks, 1, 223238 (1988).
Nilson, N.J., Learning Machines, McGraw-Hill, New York (1965).
Nishimori, H., W. Whyte, D. Sherrington, Finite-Dimensional Neural Networks Storing Structured Patterns, preprint University of Oxford (1994).
Parisi, G., Attractor Neural Networks,preprint (1994) available from cond-mat® babbage.sissa.itunder reference 941–2030.
Pastur, L.A., A.L. Figotin, Exactly Solvable Model of a Spin Glass, Soy. J. Low Temp. Phys., 3, 378–383 (1977).
Pastur, L.A., A.L. Figotin, On the Theory of Disordered Spin Systems, Theor. Math. Phys., 35, 404–414 (1978).
Pastur, L.A., M. Shcherbina, Absence of Self-Averaging of the Order Parameter in the Sherrington-Kirkpatrick Model, J. Stat. Phys., 62, 1–19 (1991).
Pastur, L.A., M. Shcherbina, B. Tirozzi, The Replica Symmetric Solution without Replica Trick for the Hopfield Model, J. Stat. Phys., 74, 1161–1183 (1994).
Patrick, A.E., V.A. Zagrebnov, Parallel Dynamics for an Extremely Diluted Neural Network, J. Phys. A: Math. Gen., 23, L1323 - L1329 (1990).
Patrick, A.E., V.A. Zagrebnov, A Probabilistic Approach to Parallel Dynamics for the Little Hopfield Model, J. Phys. A: Math. Gen., 24, 3413–3426 (1991).
Petritis, D., Simulations Numériques Monte Carlo, preprint Université de Rennes I, to be published by Masson, Paris (1995).
Petritis, D., Equilibrium Statistical Mechanics of Frustrated Disordered Systems: a Survey of Mathematical Results,preprint Université de Rennes I (1994), available from cond-matebabbage sissa. it under reference 941–1098.
Picco, P., Artificial Neural Networks, preprint CNRS Marseille (1995).
Radons, G., H.G. Schuster, D. Werner, Fractal Measures and Diffusion as Results of Learning in Neural Networks, Phys. Lett. A, 174, 293–297 (1993).
Rockafellar, R.T., Convex Analysis, Princeton Univ. Press, Princeton (1970).
Rosenblatt, F., Principles of Neurodynamics, Spartan, New York (1962).
Ruelle, D., Thermodynamic formalism, Addison-Wesley, Reading (1978).
Rumelhart, D.E., G.E. Hinton, R.J. Williams, Learning Representations by Back-Propagating Errors, Nature, 323, 533— (1986).
Sakaguchi, H., Learning Rules for an Oscillator Network, Phys. Lett. A, 174, 289–292 (1993).
Scacciatelli, E., B. Tirrozi, Fluctuation of the Free Energy in the Hopfield Model, J. Stat. Phys., 6’T, 981–1008 (1992).
Scharze, H., J. Hertz, Learning from Examples in fully Connected Committee Machines, J. Phys. A: Math. Gen., 26, 4919–4936 (1993).
Schlüter, M., E. Wagner, Phys. Rev., E49, 1690— (1994).
Shcherbina, M., B. Tirozzi, The Free Energy for for a Class of Hopfield Models, J. Stat. Phys., 72, 113–125 (1993).
Simon, B., The Statistical Mechanics of Lattice Gases, Princeton University press, Princeton (1993).
Solla, S., Learning and Generalisation in Layered Neural Networks, in Redes neuronales, teoria y aplicaciones, Escuela de Verano 88 en Física estadística y sistemas cooperativos (1988).
Solla, S., A Theory of Supervised Learning, in Neural networks: from biology to high energy physics, Proc. Elba Int. Physics Centre, O. Benhar, C. Bosio, P. del Giudice, E. Tabet eds. ETS Editrice, Pisa (1991).
Talagrand, M., Concentration of Measure and Isoperimetric Inequalities in Product Spaces, preprint Université Paris VI (1995).
Talagrand, M., A New Look at Independence, preprint Université Paris VI (1995).
Talagrand, M., Résultats Rigoureux pour le Modèle de Hopfield, preprint Université de Paris VI (1995).
Tarkowski, W., M. Lewenstein, Storage of Sets of Correlated Data in Neural Network Memories, J. Phys. A: Math. Gen., 26, 2453–2469 (1993).
Tishby, N., E. Levin, S. Solla, Consistent Inference of Probabilities in Layered Networks: Predictions and Generalisation, IEEE Neural Net., 2, 403–410 (1989).
Vermet, F., Convergence de la Variance de l’Énergie Libre du Modèle de Hopfield, C. R. Acad. Sci. Paris, 315, 1001–1004 (1992).
Vermet, F., Étude Asymptotique d’un Réseau Neuronal: le Modale de Mémoire Associative de Hopfield, PhD Thesis, Université de Rennes I, presented on 28 January (1994).
Vermet, F., Asymptotic Study of a Neural Network, preprint Université de Rennes I (1994).
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 1996 Springer Science+Business Media Dordrecht
About this chapter
Cite this chapter
Petritis, D. (1996). Thermodynamic Formalism of Neural Computing. In: Goles, E., Martínez, S. (eds) Dynamics of Complex Interacting Systems. Nonlinear Phenomena and Complex Systems, vol 2. Springer, Dordrecht. https://doi.org/10.1007/978-94-017-1323-8_3
Download citation
DOI: https://doi.org/10.1007/978-94-017-1323-8_3
Publisher Name: Springer, Dordrecht
Print ISBN: 978-90-481-4734-2
Online ISBN: 978-94-017-1323-8
eBook Packages: Springer Book Archive