Mechanisms of Winner-Take-All and Group Selection in Neuronal Spiking Networks.

Yanqing Chen

doi:10.3389/fncom.2017.00020

Abstract

A major function of central nervous systems is to discriminate different categories or types of sensory input. Neuronal networks accomplish such tasks by learning different sensory maps at several stages of neural hierarchy, such that different neurons fire selectively to reflect different internal or external patterns and states. The exact mechanisms of such map formation processes in the brain are not completely understood. Here we study the mechanism by which a simple recurrent/reentrant neuronal network accomplish group selection and discrimination to different inputs in order to generate sensory maps. We describe the conditions and mechanism of transition from a rhythmic epileptic state (in which all neurons fire synchronized and indiscriminately to any input) to a winner-take-all state in which only a subset of neurons fire for a specific input. We prove an analytic condition under which a stable bump solution and a winner-take-all state can emerge from the local recurrent excitation-inhibition interactions in a three-layer spiking network with distinct excitatory and inhibitory populations, and demonstrate the importance of surround inhibitory connection topology on the stability of dynamic patterns in spiking neural network.

Highlights

Facing with vast amount of multi-sensory information, Central Nervous System (CNS) seems to process only a small subset of those inputs at any given time, no matter whether they come from external or internal sources
To classify different types of spike dynamics for the surround inhibition network in Figure 1, for each neuron, we record the number of spikes between 2 and 3 s after the simulation had reached steady state without synaptic plasticity (STDP off)
We characterize the behavior of the network by the maximum number of spikes generated by any excitatory neuron

Summary

Introduction

Facing with vast amount of multi-sensory information, Central Nervous System (CNS) seems to process only a small subset of those inputs at any given time, no matter whether they come from external or internal sources. Winner-Take-All (WTA) network is one of such proposed mechanisms for developing feature selectivity through competition in simple recurrent networks, and it has received much attention on both theoretical and experimental grounds. The primary theoretical justification is the ability of such networks to explain how the maps, which are ubiquitous in the cerebral cortex, can arise (Kohonen, 1982; Goodhill, 2007). WTA networks can explain how a network can come to make useful distinctions between its inputs. WTA models are often employed at the behavioral level in Analytical Conditions for Winner-Take-All Neuronal Networks theoretical models of higher-level cognitive phenomenon such as action-selection, attention (Itti and Koch, 2001; Walther and Koch, 2001) and decision making (Wang, 2002; Furman and Wang, 2008)

Methods

Results

Conclusion