Generalized stochastic spiking neuron model and extended spike response model in spatial-temporal pulse pattern detection task

Authors:
O. Y. Sinyavskiy;A. I. Kobrin
Affiliations:
Moscow Power Engineering Institute (TU), Moscow, Russia 111250;Moscow Power Engineering Institute (TU), Moscow, Russia 111250
Venue:
Optical Memory and Neural Networks
Year:
2010

Citing 3
Cited 0

Role of temporal integration and fluctuation detection in the highly irregular firing of a leaky integrator neuron model with partial reset

Neural Computation
Spiking Neuron Models: An Introduction

Spiking Neuron Models: An Introduction
Optimal spike-timing-dependent plasticity for precise action potential firing in supervised learning

Neural Computation

Quantified Score

Hi-index	0.00

Visualization

Abstract

A generalized model of spiking neuron is proposed as a non-stationary stochastic spike sequences processing unit. The generalized spiking neuron model is described with a conditional spike generation probability distribution and with a state evolution operator. An information theory language is used for the convenient neuron's learning tasks description. The problem of spiking neuron learning with the teacher is solved using information entropy minimization algorithm. A particular implementation of generalized model using stochastic Spike Response Model with alpha-functions set is provided. A task of time delay maintenance between input and output spikes and a task of detecting of a spiking pattern in a noisy stream of pulse signals are considered using extended SRM neuron. It is shown that after the using of the proposed learning method spiking neuron became capable to detect a spatial-temporal pulse pattern and to serve as an adaptive delay unit.