Coupling layers regularizes wave propagation in laminar stochastic neural fields
نویسنده
چکیده
We study the effects of coupling between layers of stochastic neural field models with laminar structure. In particular, we focus on how the propagation of waves of neural activity in each layer is affected by the coupling. Synaptic connectivities within and between each layer are determined by integral kernels of an integrodifferential equation describing the temporal evolution of neural activity. Excitatory neural fields, with purely positive connectivities, support traveling fronts in each layer, whose speeds are increased when coupling between layers is considered. Studying the effects of noise, we find coupling also serves to reduce the variance in the position of traveling fronts, as long as the noise sources to each layer are not completely correlated. Neural fields with asymmetric connectivity support traveling pulses whose speeds are decreased by interlaminar coupling. Again, coupling reducers the variance in traveling pulse position, when noise is considered that is not totally correlated between layers. To derive our stochastic results, we employ a small-noise expansion, also assuming inter-laminar connectivity scales similarly. Our asymptotic results agree reasonably with accompanying numerical simulations.
منابع مشابه
Coupling layers regularizes wave propagation in stochastic neural fields.
We explore how layered architectures influence the dynamics of stochastic neural field models. Our main focus is how the propagation of waves of neural activity in each layer is affected by interlaminar coupling. Synaptic connectivities within and between each layer are determined by integral kernels of an integrodifferential equation describing the temporal evolution of neural activity. Excita...
متن کاملDelay stabilizes stochastic motion of bumps in layered neural fields
We study the effects of propagation delays on the stochastic dynamics of bumps in neural fields with multiple layers. In the absence of noise, each layer supports a stationary bump solution that can be centered at any location, due to translation invariance. Using linear stability analysis, we show that delayed coupling between layers causes translating perturbations of the bumps to decay in th...
متن کاملResponse of Two Temperatures on Wave Propagation in Micropolar Thermoelastic Materials with One Relaxation Time Bordered with Layers or Half Spaces of Inviscid Liquid
The present study is concerned with the propagation of Lamb waves in a homogeneous isotropic thermoelastic micropolar solid with two temperatures bordered with layers or half spaces of inviscid liquid subjected to stress free boundary conditions. The generalized theory of thermoelasticity developed by Lord and Shulman has been used to investigate the problem. The secular equations for symmetric...
متن کاملPropagation of Spike Sequences in Neural Networks
Precise spatiotemporal sequences of action potentials are observed in many brain areas and are thought to be involved in the neural processing of sensory stimuli. Here, we examine the ability of spiking neural networks to propagate stably a spatiotemporal sequence of spikes in the limit where each neuron fires only one spike. In contrast to previous studies on propagation in neural networks, we...
متن کاملAnalysis on the Dropout Effect in Convolutional Neural Networks
Regularizing neural networks is an important task to reduce overfitting. Dropout [1] has been a widely-used regularization trick for neural networks. In convolutional neural networks (CNNs), dropout is usually applied to the fully connected layers. Meanwhile, the regularization effect of dropout in the convolutional layers has not been thoroughly analyzed in the literature. In this paper, we an...
متن کامل