Video 9 - Spiking Neural Network Properties

This is a simulation for Spiking Neural Networks (SNN) that shows the impact of the attributes of the neuron on the output of the network in addition to the change in membrane potential and spiking patterns
As the spike reaches the output neuron its Membrane Potential rises and falls according to the Tm and Ts constants
Increasing Tm will make the voltage raise faster, while increasing Ts will make the voltage fall faster
The Axonal Delay (Delta T) contributes to the delay between receiving an input and how is this input reflected on the output
The weights of the synapses have direct impact on the Membrane Voltage
This simulation has a 1hz input
As the Membrane Potential exceeds the Voltage Threshold, the neuron emits a spike and enters a refractory period where further spikes are inhibited for a short time
One of the most important aspects of the SNN is the effect of timing on the output
The timing of the incoming spikes should be close enough to generate a Membrane Potential that exceeds the Threshold or else no spike will be emitted
Training a Spiking Neuron generally follows the Hebbian Rule where neurons that have same spiking patterns make a better bond among them
This is reflected by the changes in the weights of the synapses joining them

Category:

Science & Technology

Tags:

• University
• of
• Fukui;SNN;Spiking
• Neural
• Network;Simulation

[fifothekid]

Read the papers of floreano about spiking neural networks

[pakhtoonah]

ashkurak its helpful

[bjeddahb]

يعطيك العافية على الفديوهات...ياريت لو كان في صوت...أنا بتعلم على

SNN

إيش ممكن تنصحني أشوف أو أقرأ كمبتدئ

مهندس كمبيوتر