Summary
International Symposium on Nonlinear Theory and its Applications
2005
Session Number:2-1-3
Session:
Number:2-1-3-5
Iterative Model of Mesoscopic Neural Populations Displaying Aperiodic Dynamics
Derek Harter, Robert Kozma,
pp.602-605
Publication Date:2005/10/18
Online ISSN:2188-5079
Summary:
Mesoscopic level neurodynamics study the collective dynamical behavior of neural populations. Such models are becoming increasingly important in understanding large-scale brain processes. Mesoscopic dynamics exhibit aperiodic oscillations with a much more rich dynamical behavior than fixed-point and limit-cycle approximation allow. Far from being an undesired behavior to be suppressed, researchers are beginning to explore the idea that aperiodic dynamics may be essential to the fast recognition and large capacities of biological memories. In this paper we discuss one such mesoscopic population model, based on Freeman’s original K-set formulation. This model replicates the aperiodic behavior observed in biological brains. We are using this model to construct robot controllers that utilize such rich dynamics as their mechanisms for forming meanings and acting on past experiences. In this paper we introduce a discrete approximation of the original K-set continuous ODE model. We develop the discrete time model and compare its dynamical behavior in the so called K-III realm with the continuous ODE model. We then demonstrate its usefulness as a biologically inspired robot controller.