Neural network learning of optimal Kalman prediction and control

Although there are many neural network (NN) algorithms for prediction and for control, and although methods for optimal estimation (including filtering and prediction) and for optimal control in linear systems were provided by Kalman in 1960 (with nonlinear extensions since then), there has been, to my knowledge, no NN algorithm that learns either Kalman prediction or Kalman control (apart from the special case of stationary control). Here we show how optimal Kalman prediction and control (KPC), as well as system identification, can be learned and executed by a recurrent neural network composed of linear-response nodes, using as input only a stream of noisy measurement data. The requirements of KPC appear to impose significant constraints on the allowed NN circuitry and signal flows. The NN architecture implied by these constraints bears certain resemblances to the local-circuit architecture of mammalian cerebral cortex. We discuss these resemblances, as well as caveats that limit our current ability to draw inferences for biological function. It has been suggested that the local cortical circuit (LCC) architecture may perform core functions (as yet unknown) that underlie sensory, motor, and other cortical processing. It is reasonable to conjecture that such functions may include prediction, the estimation or inference of missing or noisy sensory data, and the goal-driven generation of control signals. The resemblances found between the KPC NN architecture and that of the LCC are consistent with this conjecture.

[1]  Rajesh P. N. Rao,et al.  Dynamic Model of Visual Recognition Predicts Neural Response Properties in the Visual Cortex , 1997, Neural Computation.

[2]  Rajesh P. N. Rao,et al.  Bayesian Inference and Attentional Modulation in the Visual Cortex Correspondence and Requests for Reprints to Rajesh , 2005 .

[3]  C. A. Gallagher,et al.  Ascending Projections of Simple and Complex Cells in Layer 6 of the Cat Striate Cortex , 1998, The Journal of Neuroscience.

[4]  J. Nazuno.  Haykin, Simon. Neural networks: A comprehensive foundation, Prentice Hall, Inc. Segunda Edición, 1999 , 2000 .

[5]  Ralph Linsker,et al.  Local Synaptic Learning Rules Suffice to Maximize Mutual Information in a Linear Network , 1992, Neural Computation.

[6]  Konrad Paul Kording,et al.  Bayesian integration in sensorimotor learning , 2004, Nature.

[7]  S. Grossberg,et al.  Context-sensitive binding by the laminar circuits of V1 and V2: A unified model of perceptual grouping, attention, and orientation contrast , 2001 .

[8]  Sharad Singhal,et al.  Training Multilayer Perceptrons with the Extende Kalman Algorithm , 1988, NIPS.

[9]  William Bialek,et al.  Spikes: Exploring the Neural Code , 1996 .

[10]  S. Hyakin,et al.  Neural Networks: A Comprehensive Foundation , 1994 .

[11]  R. Douglas,et al.  Neuronal circuits of the neocortex. , 2004, Annual review of neuroscience.

[12]  Terrence J. Sejnowski,et al.  Bayesian Unsupervised Learning of Higher Order Structure , 1996, NIPS.

[13]  Ronald J. Williams,et al.  Training recurrent networks using the extended Kalman filter , 1992, [Proceedings 1992] IJCNN International Joint Conference on Neural Networks.

[14]  E. Callaway.  Local circuits in primary visual cortex of the macaque monkey. , 1998, Annual review of neuroscience.

[15]  S. Haykin.  Kalman Filtering and Neural Networks , 2001 .

[16]  D. George,et al.  A hierarchical Bayesian model of invariant pattern recognition in the visual cortex , 2005, Proceedings. 2005 IEEE International Joint Conference on Neural Networks, 2005..

[17]  Geoffrey E. Hinton,et al.  Self-organizing neural network that discovers surfaces in random-dot stereograms , 1992, Nature.

[18]  Geoffrey E. Hinton,et al.  Generative models for discovering sparse distributed representations. , 1997, Philosophical transactions of the Royal Society of London. Series B, Biological sciences.

[19]  Tai Sing Lee,et al.  Hierarchical Bayesian inference in the visual cortex. , 2003, Journal of the Optical Society of America. A, Optics, image science, and vision.

[20]  Léon Personnaz,et al.  A recursive algorithm based on the extended Kalman filter for the training of feedforward neural models , 1998, Neurocomputing.

[21]  Ralph Linsker,et al.  Improved local learning rule for information maximization and related applications , 2005, Neural Networks.

[22]  Rajesh P. N. Rao.  Bayesian Computation in Recurrent Neural Circuits , 2004, Neural Computation.

[23]  Yee Whye Teh,et al.  A Fast Learning Algorithm for Deep Belief Nets , 2006, Neural Computation.

[24]  Rajesh P. N. Rao,et al.  An optimal estimation approach to visual perception and learning , 1999, Vision Research.

[25]  Barnabás Póczos,et al.  Neural Kalman filter , 2005, Neurocomputing.

[26]  C. Gilbert.  Microcircuitry of the visual cortex. , 1983, Annual review of neuroscience.

[27]  Franklin A. Graybill,et al.  Introduction to The theory , 1974 .

[28]  Peter Dayan,et al.  Probabilistic Computation in Spiking Populations , 2004, NIPS.

[29]  Emanuel Todorov,et al.  Stochastic Optimal Control and Estimation Methods Adapted to the Noise Characteristics of the Sensorimotor System , 2005, Neural Computation.

[30]  András Lörincz,et al.  Erratum , 2007, Neural Computation.

[31]  Anders Krogh,et al.  Introduction to the theory of neural computation , 1994, The advanced book program.

[32]  Peter Dayan,et al.  Inference, Attention, and Decision in a Bayesian Neural Architecture , 2004, NIPS.

[33]  V. Mountcastle.  Perceptual Neuroscience: The Cerebral Cortex , 1998 .

[34]  T. Başar,et al.  A New Approach to Linear Filtering and Prediction Problems , 2001 .

[35]  S. Grossberg,et al.  A neural model of how horizontal and interlaminar connections of visual cortex develop into adult circuits that carry out perceptual grouping and learning. , 2010, Cerebral cortex.

[36]  Andreas Ziehe,et al.  Adaptive On-line Learning in Changing Environments , 1996, NIPS.

[37]  Tomaso Poggio,et al.  Generalization in vision and motor control , 2004, Nature.