The Effect Of The Synaptic Weight Correction And Inhibitory Neuron Number On Brain Firing Activities

There’s several billions of neuron contained on human brain with more than1410 synapses.The research about neuron must on different aspect since neuron’s structure and function are extremely complex.Neural network and whole brain function is the emphasis on neural informatics’ research and human brain project.Synapse is the main structure for information transmits between neurons from the abilities under some certainly conditions to change its’ number and shape and to regulate its’ function,i.e.synaptic plasticity.Many researches prove that synaptic plasticity is the cell and molecular biological basic of learning and memory.We’ve studied the effect of neuron’s spiking activities from synaptic plasticity correction parameters and proportion of inhibitory neurons by neural network model.This research contains the following 3 parts:First part contains the study of effect of neural network’s spiking pattern by corrected intension of synaptic weight.In plastic neural network,we use Izhikevich model to describe the kinetics of single neuron and spike-timing-dependent-plasticity(STDP)rule for the synaptic plasticity.We also consider the conduction delay of spikes on axon.From this research we find that there are ? rhythm appeared on the spiking activities and the presence depends on the corrected intension of the synaptic weights: When the maximum value of synaptic weight incremental correction(10)A at the region 0.09~0.13,the ? rhythm appears in network’s 3600 s spiking time.When0.09 A0.1(10)? ?,With the increase of(10)A,? rhythm’s first appear will be postponed,when 0.1 A0.13(10)? ?,With the increase of(10)A,? rhythm’s first appear will in advance.When the maximum value of synaptic weight inhibited correction-A at the region 0.12~0.13,the ? rhythm appears in network’s 3600 s spiking time.With the increase of-A,? rhythm’s first appear will in advance.This part contented in the second chapter of thesis.Second part contains the study of effect of neural network’s spiking pattern by time parameter ? which describe the weight’s changing rate.We find that ?rhythm will appear in network’s 3600 s spiking time when ? at the region 7~21.With the decrease of ?,? rhythm’s first appear will in advance.This part contented in the second chapter of thesis.Third part contains the study of effect of neural network’s spiking rhythm by the proportion of inhibitory neurons.There are excitatory and inhibitory neurons in neural network since the balance of EEG.Excitatory neuron is responsible for information transmit;inhibitory neuron can make the activity of neural network on a stabilized state from releasing neurotransmitters.Inhibitory neuron is very important to neural coding in neural network although there are less inhibitory neurons than excitatory neurons.Our study obtains that the presence of ? rhythm relies on the proportion of inhibitory neuron,network’s 3600 s spiking time will contain ? rhythm if the proportion Ne:Ni satisfy ??174:826Ni:Ne200:800(Where Ne: Ni is the ratio of excitatory neurons to the number of inhibitory neurons).In addition,we study the change of ? rhythm’s first appear effected by the proportion of inhibitory neuron.There are 3 special proportions in the changing process.By contrast the time-varying number of spiking neuron in network,we find that ? rhythm will appear when there are few large quantity spiking activities.On the contrary,if there are numerous large quantity spiking activities,? rhythm will not appear in spiking frequency spectrum distribution diagram.This part contented in the third chapter of thesis.