The brain's recognition dynamics: A statistical mechanical formulation

Chang Sub Kim (Department of Physics, Chonnam National University)

The brain is an organism which perceives the external world based on uncertain sensory input, given the prior belief about the cause which is presumed to be embedded in the brain in the course of evolution as well as development [1]. The information content of the perception process must be present in the spatio-temporal patterns of the neuronal correlation occurring in the grey matter. Neurons are the building blocks of the grey matter through a collective association of which the complex phenomenon of consciousness is emergent. Currently, much scientific efforts are being made to understand the underlying principles of the brain function as a Bayesian inference machine [2] and the actual biophysical implementation of the recognition dynamics (or filtering) in the brain [3]. Previously, I reported that the theoretical framework of minimizing the informational free energy, which accounts for how the brain may operate on action, perception, and learning in a unified fashion, can be recast into Hamilton's principle in terms of the Lagrangian abstractly [4].

 I formulate here a Hamilton-Jacobi-Bellman-type equation which refines  the author's early attempt to prescribe the brain's adaptive dynamics with emphasis on its  mechanical prospect. In addition, I will discuss how the probability densities entering in the free energy principle can arise from the minimal biophysical model for the neuronal states [5], applying formal statistical physics technique [6]. In doing so, a key issue is how to incorporate the adaptiveness (or control) of a biological agent in the existing materialistic picture.


[1]  von Helmholtz H (1909) Trietise on Physiological Optics.Vol. III 3rd ed. Voss Hamburg.

[2]  Friston K (2010) The free-energy principle: a unified brain theory? Naturereivew Neurosci. 11: 127-138.

[3]  Kim C S, Seth A K (2011) The free energy principle in neuroscience: A technical evaluation. in preparation.

[4] Kim C S (2010). The adaptive dynamics of brains: Lagrangian formulation. Front. Neurosci. Conference Abstract: Neuroinformatics 2010 . doi: 10.3389/conf.fnins.2010.13.00046

[5]  Hodgkin A, Huxley A (1952) A quantitative description of membrane current and its application to conduction and excitation in nerve. J. Physiol. 117:500-544.

[6]  Zwanzig R (2001) Nonequilibrium Statistical Mechanics. Oxford University Press. New York.

Preferred presentation format: Poster
Topic: Computational neuroscience

Document Actions