首页function [y_downsampled,tau]=neuron(v_pre,delta_t,y_init,g_l,v_l,c_m,sigma,mu,W,E) y_1 = y_init; %[v_pre_resampled time_ref]= resample(v_pre,[1:size(v_pre,1)],1/delta_t); for i = 1:size(v_pre,1) dydt = (-(g_l + sum(W.*1./(1+ exp(-sigma'.*(v_pre(i,:)-mu')))',1)').*y_1 + g_l.*v_l + sum(W.*1./(1+ exp(-sigma'.*(v_pre(i,:)-mu')))'.*E,1)')./c_m; y(i,1) = y_1 + delta_t * dydt; y_1 = y(i,1); tau(i,1) =1./((g_l + sum(W.*1./(1+ exp(-sigma'.*(v_pre(i,:)-mu')))',1)')./c_m); end y_downsampled = resample(y,[1:size(v_pre,1)],1/1); %figure;plot(y) clear g_l v_l c_m sigma mu W E end

function [y_downsampled,tau]=neuron(v_pre,delta_t,y_init,g_l,v_l,c_m,sigma,mu,W,E) y_1 = y_init; %[v_pre_resampled time_ref]= resample(v_pre,[1:size(v_pre,1)],1/delta_t); for i = 1:size(v_pre,1) dydt = (-(g_l + sum(W.1./(1+ exp(-sigma'.(v_pre(i,:)-mu')))',1)').y_1 + g_l.v_l + sum(W.1./(1+ exp(-sigma'.(v_pre(i,:)-mu')))'.E,1)')./c_m; y(i,1) = y_1 + delta_t dydt; y_1 = y(i,1); tau(i,1) =1./((g_l + sum(W.1./(1+ exp(-sigma'.(v_pre(i,:)-mu')))',1)')./c_m); end y_downsampled = resample(y,[1:size(v_pre,1)],1/1); %figure;plot(y) clear g_l v_l c_m sigma mu W E end

时间: 2024-04-26 07:26:26 浏览: 47

lyaprosen.rar_The First_exp

NOTE1: When user do not have any information about tau, she should let tau equal to zero(0). In this case the code will use autocorrelation up to orders 10 to select proper embedding lag(tau). The proper lag is the lag before of first decline of autocorrelation value below exp(-1)=0.367879441.For data with nonlinear dependency autocorrelation function is not proper and mutual information criteria will be used for selecting proper lag value(tau). when both of criteria , Autocorrelation and mutual information fail to select tau, tau=1 is selected automatically

这是一个 Matlab 函数，其功能为模拟神经元的行为。具体来说，该函数接受一些输入参数（如预先记录的电压信号 v_pre，时间步长 delta_t，初始状态 y_init，电导 g_l、电压阈值 v_l、膜电容 c_m、激活函数参数 sigma、mu、连接权重矩阵 W，以及静息电位 E 等），并通过计算得出相应的输出值（如下采样后的状态 y_downsampled，时间常数 tau 等）。函数中包含了一些基本的运算，如矩阵乘法、求和、sigmoid 函数等，用于模拟神经元的膜电位变化及其对应的动作电位输出。

阅读全文

最新推荐

相关推荐

生成正态白噪声序列及其统计特性分析

Wigner-Ville分布的时频特性及其应用

解释：class DDPGAgent: def __init__(self, state_dim, action_dim, gamma=0.99, tau=0.01, lr_actor=1e-3, lr_critic=1e-3, memory_size=int(1e6), batch_size=128, warmup_steps=1000, noise_std=0.2, noise_min=0., hidden_size=128, num_layers=2)

由\begin{aligned} C_m\frac{dV}{dt} &= -g_L(V-E_L) + I_{syn}(t) 怎么能推导出\frac{d}{dt} \sum_{i=1}^{N} w_i S_i(t) &= -\sum_{i=1}^{N} \frac{S_i(t)}{\tau_i} + \sum_{k=1}^{K} \sum_{j=1}^{N_k} \delta(t-t_j^{(k)})w_{ij} \

$$ \begin{aligned} C_m\frac{dV}{dt} &= -g_L(V-E_L) + I_{syn}(t) \ \frac{d}{dt} \sum_{i=1}^{N} w_i S_i(t) &= -\sum_{i=1}^{N} \frac{S_i(t)}{\tau_i} + \sum_{k=1}^{K} \sum_{j=1}^{N_k} \delta(t-t_j^{(k)})w_{ij} \ \end{aligned} $$ 是什么意思

解释下这串代码：R=zeros(2*max_tau+1,L); for tau=-max_tau:max_tau x1=x; x2=wshift(1, x, tau); R(tau+1+max_tau,:)=exp(-1i*pi*((0:L-1)/L)*tau).*fft(para.*exp(-(x1-x2).^2/(2*kernelsize^2))/L);%

在$$ U（1）_ {L_ \ mu-L_ \ tau} $$ U（1）Lμ-Lτ模型的对撞机实验中搜索标量玻色子衰减为光$$ Z'$$ Z'玻色子

for i = 1:size(v_pre,1) dydt = (-(g_l + sum(W.*1./(1+ exp(-sigma'.*(v_pre(i,:)-mu')))',1)').*y_1 + g_l.*v_l + sum(W.*1./(1+ exp(-sigma'.*(v_pre(i,:)-mu')))'.*E,1)')./c_m; y(i,1) = y_1 + delta_t * dydt; y_1 = y(i,1); tau(i,1) =1./((g_l + sum(W.*1./(1+ exp(-sigma'.*(v_pre(i,:)-mu')))',1)')./c_m); end

解释function [X, A, B, S, Out] = Mixed_LRTF_DFR(Y, opts) max_it = opts.max_it; Bmax_it = opts.Bmax_it; tol = opts.tol; R = opts.R; rho = opts.rho; tau = opts.tau; lambda = opts.lambda; beta = opts.beta; mu = opts.mu; Out.Res=[]; Out.PSNR=[];

multistep_preds = torch.zeros(T) multistep_preds[: n_train + tau] = x[: n_train + tau] for i in range(n_train + tau, T): multistep_preds[i] = net( multistep_preds[i - tau:i].reshape((1, -1)))的作用

一阶系统参数辨识：最小二乘法应用解析

最新推荐

NR5G网络拒绝码-5gmm_cause = 7 (0x7) (5GS Service not allowed)

vb.net+sql毕业设计管理系统(论文+系统)(2024hu).7z

RStudio中集成Connections包以优化数据库连接管理

管理建模和仿真的文件

Keil uVision5全面精通指南

flink提交给yarn19个全量同步MYsqlCDC的作业，flink的配置参数怎样设置

PHP博客旅游的探索之旅

"互动学习：行动中的多样性与论文攻读经历"

【单片机编程实战】：掌握流水灯与音乐盒同步控制的高级技巧

java 号码后四位用‘xxxx’脱敏

解释：class DDPGAgent: def init(self, state_dim, action_dim, gamma=0.99, tau=0.01, lr_actor=1e-3, lr_critic=1e-3, memory_size=int(1e6), batch_size=128, warmup_steps=1000, noise_std=0.2, noise_min=0., hidden_size=128, num_layers=2)

解释下这串代码：R=zeros(2max_tau+1,L); for tau=-max_tau:max_tau x1=x; x2=wshift(1, x, tau); R(tau+1+max_tau,:)=exp(-1ipi((0:L-1)/L)tau).fft(para.exp(-(x1-x2).^2/(2*kernelsize^2))/L);%

for i = 1:size(v_pre,1) dydt = (-(g_l + sum(W.1./(1+ exp(-sigma'.(v_pre(i,:)-mu')))',1)').y_1 + g_l.v_l + sum(W.1./(1+ exp(-sigma'.(v_pre(i,:)-mu')))'.E,1)')./c_m; y(i,1) = y_1 + delta_t dydt; y_1 = y(i,1); tau(i,1) =1./((g_l + sum(W.1./(1+ exp(-sigma'.(v_pre(i,:)-mu')))',1)')./c_m); end