Building fast Bayesian computing machines out of intentionally stochastic, digital parts

The brain interprets ambiguous sensory information faster and more reliably than modern computers, using neurons that are slower and less reliable than logic gates. But Bayesian inference, which underpins many computational models of perception and cognition, appears computationally challenging even given modern transistor speeds and energy budgets. The computational principles and structures needed to narrow this gap are unknown. Here we show how to build fast Bayesian computing machines using intentionally stochastic, digital parts, narrowing this efficiency gap by multiple orders of magnitude. We find that by connecting stochastic digital components according to simple mathematical rules, one can build massively parallel, low precision circuits that solve Bayesian inference problems and are compatible with the Poisson firing statistics of cortical neurons. We evaluate circuits for depth and motion perception, perceptual learning and causal reasoning, each performing inference over 10,000+ latent variables in real time - a 1,000x speed advantage over commodity microprocessors. These results suggest a new role for randomness in the engineering and reverse-engineering of intelligent computation.

[1]  Persi Diaconis,et al.  The Markov chain Monte Carlo revolution , 2008 .

[2]  V.D. Agrawal,et al.  Soft Error Rate Determination for Nanometer CMOS VLSI Logic , 2008, 2008 40th Southeastern Symposium on System Theory (SSST).

[3]  Neil Gershenfeld,et al.  Continuous-time analog circuits for statistical signal processing , 2003 .

[4]  Michael R. Waldmann,et al.  Causal Reasoning in Rats , 2006, Science.

[5]  Joshua B. Tenenbaum,et al.  Natively probabilistic computation , 2009 .

[6]  Kwabena Boahen,et al.  Silicon Neurons That Compute , 2012, ICANN.

[7]  Christian Rosenmund,et al.  Nonuniform probability of glutamate release at a hippocampal synapse. , 1993, Science.

[8]  Geoffrey E. Hinton,et al.  Deep Boltzmann Machines , 2009, AISTATS.

[9]  Adam N. Sanborn,et al.  Categorization as nonparametric Bayesian density estimation , 2008 .

[10]  Andrew W. Fitzgibbon,et al.  Real-time human pose recognition in parts from single depth images , 2011, CVPR 2011.

[11]  Charles Kemp,et al.  How to Grow a Mind: Statistics, Structure, and Abstraction , 2011, Science.

[12]  József Fiser,et al.  Spontaneous Cortical Activity Reveals Hallmarks of an Optimal Internal Model of the Environment , 2011, Science.

[13]  N. Metropolis,et al.  Equation of State Calculations by Fast Computing Machines , 1953, Resonance.

[14]  J. Flinn,et al.  Energy-aware adaptation for mobile applications , 2000, OPSR.

[15]  M. Elowitz,et al.  A synthetic oscillatory network of transcriptional regulators , 2000, Nature.

[16]  Sebastian Thrun,et al.  Probabilistic robotics , 2002, CACM.

[17]  Claude E. Shannon,et al.  A symbolic analysis of relay and switching circuits , 1938, Transactions of the American Institute of Electrical Engineers.

[18]  Nando de Freitas,et al.  An Introduction to MCMC for Machine Learning , 2004, Machine Learning.

[19]  A. Hasman,et al.  Probabilistic reasoning in intelligent systems: Networks of plausible inference , 1991 .

[20]  A Imre,et al.  Majority Logic Gate for Magnetic Quantum-Dot Cellular Automata , 2006, Science.

[21]  Judea Pearl,et al.  Probabilistic reasoning in intelligent systems - networks of plausible inference , 1991, Morgan Kaufmann series in representation and reasoning.

[22]  Brendan J. Frey,et al.  Factor graphs and the sum-product algorithm , 2001, IEEE Trans. Inf. Theory.

[23]  Kenneth L. Shepard,et al.  Noise in deep submicron digital design , 1996, Proceedings of International Conference on Computer Aided Design.

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

[25]  D Marr,et al.  Cooperative computation of stereo disparity. , 1976, Science.

[26]  Donald Geman,et al.  Stochastic Relaxation, Gibbs Distributions, and the Bayesian Restoration of Images , 1984, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[27]  A. Arkin,et al.  Stochastic mechanisms in gene expression. , 1997, Proceedings of the National Academy of Sciences of the United States of America.

[28]  Edward H. Adelson,et al.  Motion illusions as optimal percepts , 2002, Nature Neuroscience.

[29]  Brian R. Gaines,et al.  Stochastic Computing Systems , 1969 .

[30]  Robert H. Halstead,et al.  Computation structures , 1990, MIT electrical engineering and computer science series.

[31]  Jean,et al.  The Computer and the Brain , 1989, Annals of the History of Computing.

[32]  John R. Anderson,et al.  A Rational Analysis of Categorization , 1990, ML.

[33]  T. Ferguson A Bayesian Analysis of Some Nonparametric Problems , 1973 .

[34]  Richard Szeliski,et al.  A Comparative Study of Energy Minimization Methods for Markov Random Fields with Smoothness-Based Priors , 2008, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[35]  J. C. McPherson Papers and discussions presented at the Dec. 10-12, 1951, joint AIEE-IRE computer conference: Review of electronic digital computers , 1951 .

[36]  J. Presper Eckert,et al.  The UNIVAC system , 1951, AIEE-IRE '51.

[37]  John Wawrzynek,et al.  High-throughput bayesian computing machine with reconfigurable hardware , 2010, FPGA '10.

[38]  A. Pouget,et al.  Probabilistic brains: knowns and unknowns , 2013, Nature Neuroscience.

[39]  Michelle Becker The Probabilistic Mind Prospects For Bayesian Cognitive Science , 2016 .

[40]  Yann LeCun,et al.  The mnist database of handwritten digits , 2005 .

[41]  J. Tenenbaum,et al.  Optimal Predictions in Everyday Cognition , 2006, Psychological science.

[42]  Lorenzo Alvisi,et al.  Modeling the effect of technology trends on the soft error rate of combinational logic , 2002, Proceedings International Conference on Dependable Systems and Networks.

[43]  Krishna V. Palem,et al.  Probabilistic CMOS Technology: A Survey and Future Directions , 2006, 2006 IFIP International Conference on Very Large Scale Integration.

[44]  Carver A. Mead,et al.  Neuromorphic electronic systems , 1990, Proc. IEEE.

[45]  Jennifer Chu-Carroll,et al.  Building Watson: An Overview of the DeepQA Project , 2010, AI Mag..

[46]  Luiz André Barroso,et al.  The Case for Energy-Proportional Computing , 2007, Computer.

[47]  Carl E. Rasmussen,et al.  The Infinite Gaussian Mixture Model , 1999, NIPS.

[48]  X. Jin Factor graphs and the Sum-Product Algorithm , 2002 .

[49]  Michael Luby,et al.  An Optimal Approximation Algorithm for Bayesian Inference , 1997, Artif. Intell..

[50]  P. Berkes,et al.  Statistically Optimal Perception and Learning: from Behavior to Neural Representations , 2022 .

[51]  Joel S. Emer,et al.  Techniques to reduce the soft error rate of a high-performance microprocessor , 2004, Proceedings. 31st Annual International Symposium on Computer Architecture, 2004..