Evolutionary supervision of a dynamical neural network allows learning with on-going weights

Recent electrophysiological data show that synaptic weights are highly influenced by electrical activities displayed by neurons. Weights are not stable as assumed in classical neural network models. What is the nature of engrains, if not stored in synaptic weights? Adopting the theory of dynamical systems, which allows an implicit form of memory, we propose a new framework for learning, where synaptic weights are continuously adapted. Evolutionary computation has been applied to a population of dynamic neural networks evolving in a prey-predator environment. Each individual develops complex dynamic patterns of neuronal activity, underlied by multiple recurrent connections. We show that this method allows the emergence of learning capability through generations, as a byproduct of evolution, since the behavioural performance of the network is not a priori based on this property.

[1]  R. French.  Catastrophic Forgetting in Connectionist Networks , 2006 .

[2]  G. Edelman,et al.  Neural dynamics in a model of the thalamocortical system. I. Layers, loops and the emergence of fast synchronous rhythms. , 1997, Cerebral cortex.

[3]  Geoffrey E. Hinton,et al.  How Learning Can Guide Evolution , 1996, Complex Syst..

[4]  Kunihiko Kaneko,et al.  ISSUE : Chaotic Itinerancy Chaotic itinerancy , 2003 .

[5]  G. Edelman.  Neural Darwinism: The Theory Of Neuronal Group Selection , 1989 .

[6]  J. Baldwin.  A New Factor in Evolution , 1896, The American Naturalist.

[7]  Didier Puzenat,et al.  A multisensory identification system for robotics , 2001, IJCNN'01. International Joint Conference on Neural Networks. Proceedings (Cat. No.01CH37222).

[8]  I. Tsuda.  Toward an interpretation of dynamic neural activity in terms of chaotic dynamical systems. , 2001, The Behavioral and brain sciences.

[9]  G. Laurent,et al.  Odor encoding as an active, dynamical process: experiments, computation, and theory. , 2001, Annual review of neuroscience.

[10]  T. Gelder,et al.  The dynamical hypothesis in cognitive science , 1998, Behavioral and Brain Sciences.

[11]  Thomas Nowotny,et al.  Enhancement of Synchronization in a Hybrid Neural Circuit by Spike-Timing Dependent Plasticity , 2003, The Journal of Neuroscience.

[12]  Christoph von der Malsburg,et al.  The What and Why of Binding The Modeler’s Perspective , 1999, Neuron.

[13]  G. Simpson.  THE BALDWIN EFFECT , 1953 .

[14]  Tariq Samad,et al.  Towards the Genetic Synthesisof Neural Networks , 1989, ICGA.

[15]  Bernard Ans,et al.  Neural networks with a self-refreshing memory: Knowledge transfer in sequential learning tasks without catastrophic forgetting , 2000, Connect. Sci..

[16]  Wulfram Gerstner,et al.  Why spikes? Hebbian learning and retrieval of time-resolved excitation patterns , 1993, Biological Cybernetics.

[17]  M. Poo,et al.  Coincident Pre- and Postsynaptic Activity Modifies GABAergic Synapses by Postsynaptic Changes in Cl− Transporter Activity , 2003, Neuron.

[18]  C Koch,et al.  Complexity and the nervous system. , 1999, Science.

[19]  W. Singer,et al.  Stimulus-specific neuronal oscillations in orientation columns of cat visual cortex. , 1989, Proceedings of the National Academy of Sciences of the United States of America.

[20]  G. Bi,et al.  Synaptic Modifications in Cultured Hippocampal Neurons: Dependence on Spike Timing, Synaptic Strength, and Postsynaptic Cell Type , 1998, The Journal of Neuroscience.

[21]  Giles Mayley.  Guiding or Hiding: Explorations into the Effects of Learning on the Rate of Evolution , 1997 .

[22]  L. Abbott,et al.  Synaptic plasticity: taming the beast , 2000, Nature Neuroscience.

[23]  A. Lw,et al.  A Quantitative Model of the Simpson – Baldwin Effect , 1998 .

[24]  M Kaufman,et al.  Positive feedback circuits and memory. , 2000, Comptes rendus de l'Academie des sciences. Serie III, Sciences de la vie.

[25]  H. Markram,et al.  Regulation of Synaptic Efficacy by Coincidence of Postsynaptic APs and EPSPs , 1997, Science.

[26]  Christoph von der Malsburg,et al.  The Correlation Theory of Brain Function , 1994 .

[27]  Alex M. Andrew,et al.  Spiking Neuron Models: Single Neurons, Populations, Plasticity , 2003 .

[28]  David E. Goldberg,et al.  Genetic Algorithms in Search Optimization and Machine Learning , 1988 .

[29]  Leslie M Kay,et al.  A challenge to chaotic itinerancy from brain dynamics. , 2003, Chaos.

[30]  W. Freeman,et al.  How brains make chaos in order to make sense of the world , 1987, Behavioral and Brain Sciences.

[31]  Wofgang Maas,et al.  Networks of spiking neurons: the third generation of neural network models , 1997 .

[32]  G. Edelman,et al.  Spike-timing dynamics of neuronal groups. , 2004, Cerebral cortex.

[33]  R. Thomas,et al.  Multistationarity, the basis of cell differentiation and memory. I. Structural conditions of multistationarity and other nontrivial behavior. , 2001, Chaos.

[34]  R. F. Thompson,et al.  The search for the engram. , 1976, The American psychologist.

[35]  Wulfram Gerstner,et al.  A neuronal learning rule for sub-millisecond temporal coding , 1996, Nature.

[36]  P König,et al.  Direct physiological evidence for scene segmentation by temporal coding. , 1991, Proceedings of the National Academy of Sciences of the United States of America.

[37]  Sylvain Chevallier,et al.  Distributed Processing for Modelling Real-Time Multimodal Perception in a Virtual Robot , 2005, Parallel and Distributed Computing and Networks.

[38]  W. Freeman,et al.  Role of chaotic dynamics in neural plasticity. , 1994, Progress in brain research.

[39]  S. Grossberg,et al.  Pattern Recognition by Self-Organizing Neural Networks , 1991 .

[40]  Y. Dan,et al.  Spike Timing-Dependent Plasticity of Neural Circuits , 2004, Neuron.

[41]  S. Thorpe,et al.  Speed of processing in the human visual system , 1996, Nature.

[42]  Kunihiko Kaneko,et al.  Self-organized hierarchical structure in a plastic network of chaotic units , 2000, Neural Networks.

[43]  Péter Érdi,et al.  The KIV model - nonlinear spatio-temporal dynamics of the primordial vertebrate forebrain , 2003, Neurocomputing.

[44]  C. Gray.  The Temporal Correlation Hypothesis of Visual Feature Integration Still Alive and Well , 1999, Neuron.