耦合CBEM的脉冲神经元模型建模与仿真  

Modeling and simulation of coupled CBEM spiking neuron model

作　　者：张嘉瑛 贺兴时[1] 于青林 ZHANG Jiaying;HE Xingshi;YU Qinglin(School of Science,Xi’an Polytechnic University,Xi’an 710048,China;Department of Mathematics and Statistics, Thompson River University, Kamloops, V2C 0C8 Canada)

机构地区：[1]西安工程大学理学院,陕西西安710048 [2]汤普森河大学数学与统计学系,加拿大甘露市V2C0C8

出　　处：《西安工程大学学报》2021年第5期123-130,共8页Journal of Xi’an Polytechnic University

基　　金：陕西省教育厅自然科学专项(19JK0373)。

摘　　要：CBEM脉冲神经元模型在神经编码的统计和生物物理模型之间建立了一个新的联系。为了缩小CBEM与生物物理模型之间的差距,提出了耦合CBEM的脉冲神经元模型。通过使用耦合滤波器对CBEM与脉冲神经元模型间进行耦合,针对不同脉冲序列间的相关关系,获得更多的信息并进行序列拟合。通过对模型可视化,计算单脉冲信息与AIC值。结果表明:提出的耦合CBEM在保留生物物理可解释性的基础上,获得了更加接近真实的脉冲序列仿真结果,可传输更多的脉冲序列信息,且模型的复杂度较低。The CBEM spiking neuron model establishes a new connection between the statistics of neural coding and the biophysical model.In order to narrow the gap between the CBEM model and the biophysical model,a coupled CBEM spike neuron model is proposed.A coupling filter was used to couple the single CBEM model.The correlation between different spiking sequences is considered in the coupled CBEM spiking neuron model,and the correlation information is used for sequence fitting.Through the spike rate visualization,Single-spike information and the AIC criterion show that the proposed coupled CBEM model can obtain more real spiking sequences simulation results,more spiking sequences information and the low model complexity while retaining the biophysical interpretability.

关 键 词：耦合CBEM 神经元模型 单脉冲信息 耦合滤波器 脉冲序列信息 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]