Synchronization of heteroclinic circuits through learning in coupled neural networks

详细信息    查看全文

• 作者：Anton Selskii ; Valeri A. Makarov
• 关键词：34C15 ; 37C29 ; 92B20 ; 92B25 ; synchronization ; learning ; heteroclinic circuit ; neural networks ; winner ; less competition
• 刊名：Regular and Chaotic Dynamics
• 出版年：2016
• 出版时间：January 2016
• 年：2016
• 卷：21
• 期：1
• 页码：97-106
• 全文大小：1,259 KB
• 参考文献：1.Ashwin, P. and Chossat, P., Attractors for Robust Heteroclinic Cycles with Continua of Connections, J. Nonlinear Sci., 1998, vol. 8, no. 2, pp. 103–129.CrossRef MathSciNet MATH
  2.Ashwin, P. and Field, M., Heteroclinic Network in Coupled Cell Systems, Arch. Ration. Mech. Anal., 1999, vol. 148, no. 2, pp. 107–143.CrossRef MathSciNet MATH
  3.Cohen, M. A. and Grossberg, S., Absolute Stability of Global Pattern Formation and Parallel Memory Storage by Competitive Neural Networks, IEEE Trans. Systems Man Cybernet., 1983, vol. 13, no. 5, pp. 815–826.CrossRef MathSciNet MATH
  4.Rabinovich, M. I., Volkovskii, A., Lecanda, P., Huerta, R., Abarbanel, H. D. I., and Laurent, G., Dynamical Encoding by Networks of Competing Neuron Groups: Winnerless Competition, Phys. Rev. Lett., 2001, vol. 87, no. 6, 068102, 4 pp.CrossRef
  5.Varona, P., Rabinovich, M. I., Selverston, A. I., and Arshavsky, Y. I., Winnerless Competition between Sensory Neurons Generates Chaos: A Possible Mechanism for Molluscan Hunting Behavior, Chaos, 2002, vol. 12, no. 3, pp. 672–677.CrossRef MathSciNet MATH
  6.Rabinovich, M. I., Muezzinoglu, M. K., Strigo, I., and Bystritsky, A., Dynamical Principles of Emotion-Cognition Interaction: Mathematical Images of Mental Disorders, PLoS ONE, 2010, vol. 5, no. 9, e12547.CrossRef
  7.Bystritsky, A., Nierenberg, A.A., Feusner, J. D., and Rabinovich, M., Computational Non-Linear Dynamical Psychiatry: A New Methodological Paradigm for Diagnosis and Course of Illness, J. Psychiatr. Res., 2012, vol. 46, no. 4, pp. 428–435.CrossRef
  8.Rabinovich, M. I., Varona, P., Tristan, I., and Afraimovich, V. S., Chunking Dynamics: Heteroclinics in Mind, Front. Comput. Neurosci., 2014, vol. 8, no. 22.
  9.Afraimovich, V., Young, T.R, and Rabinovich, M. I., Hierarchical Heteroclinics in Dynamical Model of Cognitive Processes: Chunking, Internat. J. Bifur. Chaos Appl. Sci. Engrg., 2014, vol. 24, no. 10, 1450132, 15 pp.CrossRef MathSciNet
  10.Bick, C. and Rabinovich, M. I., Dynamical Origin of the Effective Storage Capacity in the Brain’s Working Memory, Phys. Rev. Lett., 2009, vol. 103, no. 21, 218101, 4 pp.CrossRef
  11.Afraimovich, V. S., Rabinovich, M. I., and Varona, P., Heteroclinic Contours in Neural Ensembles and the Winnerless Competition Principle, Internat. J. Bifur. Chaos Appl. Sci. Engrg., 2004, vol. 14, no. 4, pp. 1195–1208.CrossRef MathSciNet MATH
  12.Huerta, R. and Rabinovich, M., Reproducible Sequence Generation in Random Neural Ensembles, Phys. Rev. Lett., 2004, vol. 93, no. 23, 238104, 4 pp.CrossRef
  13.Ashwin, P. and Borresen, J., Discrete Computation Using a Perturbed Heteroclinic Network, Phys. Lett. A, 2005, vol. 347, nos. 4–6, pp. 208–214.CrossRef MathSciNet MATH
  14.Rabinovich, M. I. and Muezzinoglu, M. K., Nonlinear Dynamics of the Brain: Emotion and Cognition, Phys. Uspekhi, 2010, vol. 53, no. 4, pp. 357–372; see also: Uspekhi Fiz. Nauk, 2010, vol. 180, no. 4, pp. 371–387.CrossRef
  15.Levanova, T.A., Komarov, M. A., and Osipov, G.V., Sequential Activity and Multistability in an Ensemble of Coupled Van der Pol Oscillators, Eur. Phys. J. Special Topics, 2013, vol. 222, no. 10, pp. 2417–2428.CrossRef
  16.Abarbanel, H. D. I., Rabinovich, M. I., Selverstone, A., Bazhenov, M.V., Huerta, R., Sushchik, M. M., and Rubchinskii, L. L., Synchronization in Neural Networks, Phys. Uspekhi, 1996, vol. 39, no. 4, pp. 337–362; see also: Uspekhi Fiz. Nauk, 1996, vol. 166, no. 4, pp. 363–390.CrossRef
  17.Schulz, P.E. and Fitzgibbons, J. C., Differing Mechanisms of Expression for Short- and Long-Term Potentiation, J. Neurophysiol., 1997, vol. 78, no. 1, pp. 321–334.
  18.Fernández-Ruiz, A., Makarov, V. A., and Herreras, O., Sustained Increase of Spontaneous Input and Spike Transfer in the CA3-CA1 Pathway Following Long-Term Potentiation in vivo, Front. Neural Circuits, 2012, vol. 6, no. 71.
  19.Villacorta-Atienza, J. A. and Makarov, V.A., Neural Network Architecture for Cognitive Navigation in Dynamic Environments, IEEE Trans. Neural Netw. Learn. Syst., 2013, vol. 24, no. 12, pp. 2075–2087.CrossRef
  20.Makarov, V. A., Song, Y., Velarde, M.G., H¨ ubner, D., and Cruse, H., Elements for a General Memory Structure: Properties of Recurrent Neural Networks Used to Form Situation Models, Biol. Cybern., 2008, vol. 98, no. 5, pp. 371–395.CrossRef MATH
  
• 作者单位：Anton Selskii (1)
  Valeri A. Makarov (1) (2)
  
  1. N.I. Lobachevsky State University of Nizhny Novgorod, ul. Gagarina 23, Nizhny Novgorod, 603950, Russia
  2. Instituto de Matemática Interdiciplinar, F. CC. Matemáticas, Universidad Complutense de Madrid, Madrid, 28040, Spain
  
• 刊物类别：Mathematics and Statistics
• 刊物主题：Mathematics
  Dynamical Systems and Ergodic Theory
  Russian Library of Science
  
• 出版者：MAIK Nauka/Interperiodica distributed exclusively by Springer Science+Business Media LLC.
• ISSN：1468-4845

文摘

The synchronization of oscillatory activity in neural networks is usually implemented by coupling the state variables describing neuronal dynamics. Here we study another, but complementary mechanism based on a learning process with memory. A driver network, acting as a teacher, exhibits winner-less competition (WLC) dynamics, while a driven network, a learner, tunes its internal couplings according to the oscillations observed in the teacher. We show that under appropriate training the learner can “copy” the coupling structure and thus synchronize oscillations with the teacher. The replication of the WLC dynamics occurs for intermediate memory lengths only, consequently, the learner network exhibits a phenomenon of learning resonance. Keywords synchronization learning heteroclinic circuit neural networks winner-less competition