Advertisement

Minimizing Relative Entropy in Hierarchical Predictive Coding

  • Johan Kwisthout
Part of the Lecture Notes in Computer Science book series (LNCS, volume 8754)

Abstract

The recent Hierarchical Predictive Coding theory is a very influential theory in neuroscience that postulates that the brain continuously makes (Bayesian) predictions about sensory inputs using a generative model. The Bayesian inferences (making predictions about sensory states, estimating errors between prediction and observation, and lowering the prediction error by revising hypotheses) are assumed to allow for efficient approximate inferences in the brain. We investigate this assumption by making the conceptual ideas of how the brain may minimize prediction error computationally precise and by studying the computational complexity of these computational problems. We show that each problem is intractable in general and discuss the parameterized complexity of the problems.

Keywords

Prediction Error Posterior Distribution Bayesian Network Turing Machine Relative Entropy 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Abdelbar, A.M., Hedetniemi, S.M.: Approximating MAPs for belief networks is NP-hard and other theorems. Artificial Intelligence 102, 21–38 (1998)MathSciNetCrossRefzbMATHGoogle Scholar
  2. 2.
    Batu, T., Dasgupta, S., Kumar, R., Rubinfeld, R.: The complexity of approximating the entropy. SIAM Journal on Computing 35(1), 132–150 (2005)MathSciNetCrossRefzbMATHGoogle Scholar
  3. 3.
    Blokpoel, M., Kwisthout, J., van Rooij, I.: When can predictive brains be truly Bayesian? Frontiers in Theoretical and Philosophical Psychology 3, 406 (2012)Google Scholar
  4. 4.
    Castillo, E., Gutiérrez, J.M., Hadi, A.S.: Sensitivity analysis in discrete Bayesian networks. IEEE Transactions on Systems, Man, and Cybernetics 27, 412–423 (1997)CrossRefGoogle Scholar
  5. 5.
    Chan, H., Darwiche, A.: Sensitivity analysis in Bayesian networks: From single to multiple parameters. In: Proceedings of the 20th Conference in Uncertainty in Artificial Intelligence, pp. 67–75 (2004)Google Scholar
  6. 6.
    Clark, A.: The many faces of precision (Replies to commentaries on “Whatever next? Neural prediction, situated agents, and the future of cognitive science”). Frontiers in Theoretical and Philosophical Psychology 4, e270 (2013)Google Scholar
  7. 7.
    Clark, A.: Whatever next? Predictive brains, situated agents, and the future of cognitive science. Behavioral and Brain Sciences 36(3), 181–204 (2013)CrossRefGoogle Scholar
  8. 8.
    Cooper, G.F., Herskovitz, E.: Determination of the entropy of a belief network is NP-hard. Technical Report KSL-90-21, Stanford University. Computer Science Deptu. Knowledge Systems Laboratory (March 1990)Google Scholar
  9. 9.
    Dagum, P., Luby, M.: Approximating probabilistic inference in Bayesian belief networks is NP-hard. Artificial Intelligence 60(1), 141–153 (1993)MathSciNetCrossRefzbMATHGoogle Scholar
  10. 10.
    Darwiche, A.: Modeling and Reasoning with Bayesian Networks. CU Press, Cambridge (2009)Google Scholar
  11. 11.
    Darwiche, A., Choi, A.: Same-decision probability: A confidence measure for threshold-based decisions under noisy sensors. In: 5th European Workshop on Probabilistic Graphical Models (2010)Google Scholar
  12. 12.
    Downey, R.G., Fellows, M.R.: Parameterized Complexity. Springer, Berlin (1999)CrossRefGoogle Scholar
  13. 13.
    Friston, K.J.: The free-energy principle: A rough guide to the brain? Trends in Cognitive Sciences 13(7), 293–301 (2009)CrossRefGoogle Scholar
  14. 14.
    Friston, K.J.: The free-energy principle: A unified brain theory? Nature Reviews Neuroscience 11(2), 127–138 (2010)CrossRefGoogle Scholar
  15. 15.
    Friston, K.J., Daunizeau, J., Kilner, J., Kiebel, S.J.: Action and behavior: A free-energy formulation. Biological Cybernetics 102(3), 227–260 (2010)CrossRefGoogle Scholar
  16. 16.
    Griffiths, T.L., Chater, N., Kemp, C., Perfors, A., Tenenbaum, J.B.: Probabilistic models of cognition: Exploring representations and inductive biases. Trends in Cognitive Sciences 14(8), 357–364 (2010)CrossRefGoogle Scholar
  17. 17.
    Hohwy, J.: The Predictive Mind. Oxford University Press (2013)Google Scholar
  18. 18.
    Kilner, J.M., Friston, K.J., Frith, C.D.: The mirror-neuron system: A Bayesian perspective. Neuroreport 18, 619–623 (2007)CrossRefGoogle Scholar
  19. 19.
    Knill, D., Pouget, A.: The Bayesian brain: The role of uncertainty in neural coding and computation. Trends in Neuroscience 27(12), 712–719 (2004)CrossRefGoogle Scholar
  20. 20.
    Kwisthout, J.: The Computational Complexity of Probabilistic Networks. PhD thesis, Faculty of Science, Utrecht University, The Netherlands (2009)Google Scholar
  21. 21.
    Kwisthout, J., van der Gaag, L.C.: The computational complexity of sensitivity analysis and parameter tuning. In: Chickering, D.M., Halpern, J.Y. (eds.) Proceedings of the 24th Conference on Uncertainty in Artificial Intelligence, pp. 349–356. AUAI Press (2008)Google Scholar
  22. 22.
    Kwisthout, J., Van Rooij, I.: Predictive coding: Intractability hurdles that are yet to overcome [abstract]. In: Knauff, M., Pauen, M., Sebanz, N., Wachsmuth, I. (eds.) Proceedings of the 35th Annual Conference of the Cognitive Science Society. Cognitive Science Society, Austin (2013)Google Scholar
  23. 23.
    Park, J.-C., Lim, J.H., Choi, H., Kim, D.-S.: Predictive coding strategies for developmental neurorobotics. Frontiers in Psychology 3, 134 (2012)CrossRefGoogle Scholar
  24. 24.
    Park, J.D., Darwiche, A.: Complexity results and approximation settings for MAP explanations. Journal of Artificial Intelligence Research 21, 101–133 (2004)MathSciNetzbMATHGoogle Scholar
  25. 25.
    Saygin, A.P., Chaminade, T., Ishiguro, H., Driver, J., Frith, C.: The thing that should not be: Predictive coding and the uncanny valley in perceiving human and humanoid robot actions. Social Cognitive and Affective Neuroscience 7(4), 413–422 (2012)CrossRefGoogle Scholar
  26. 26.
    Seth, A.K., Suzuki, K., Critchley, H.D.: An interoceptive predictive coding model of conscious presence. Frontiers in Psychology 2, e395 (2011)Google Scholar
  27. 27.
    van der Gaag, L.C., Bodlaender, H.L., Feelders, A.J.: Monotonicity in Bayesian networks. In: Chickering, M., Halpern, J. (eds.) Proceedings of the Twentieth Conference on Uncertainty in Artificial Intelligence, pp. 569–576. AUAI Press, Arlington (2004)Google Scholar

Copyright information

© Springer International Publishing Switzerland 2014

Authors and Affiliations

  • Johan Kwisthout
    • 1
  1. 1.Donders Institute for Brain, Cognition and BehaviourRadboud University NijmegenNijmegenThe Netherlands

Personalised recommendations