关于尖峰神经网络的误差反向传播学习算法的一点评论✩

摘要

在用于脉冲神经网络的误差反向传播学习算法中，必须将触发时间tαt^\alphatα区分为状态函数x(t)x(t)x(t)的函数。但是这种区分是不可能直接进行的，因为不能以标准形式将其表达为x(t)x(t)x(t)的函数。为了克服这个困难，Bohte等假设脉冲发时间tαt^\alphatα和状态x(t)x(t)x(t)在t=tαt=t^αt=tα附近存在线性关系。根据此假设，泛函的Frechet导数等于可以直接轻松计算的普通函数的导数。我们在此简短说明中所做的贡献是证明，在没有线性假设的帮助下，这种微分相等实际上在数学上是正确的。

脉冲感知器

为了简洁起见，让我们考虑一个尖峰神经网络的简单构建块，即具有n个突触前神经元和一个突触后神经元的前馈尖峰感知器。

A remark on the error-backpropagation learning algorithm for spiking neural networks相关推荐

Spatio-Temporal Backpropagation for Training High-performance Spiking Neural Networks笔记
<Spatio-Temporal Backpropagation for Training High-performance Spiking Neural Networks>笔记 ABST ...
(2018)All-optical machine learning using diffractive deep neural networks
"All-optical machine learning using diffractive deep neural networks",这篇Science上的文章发表于2018 ...
论文精翻《Progressive Tandem Learning for Pattern Recognition With Deep Spiking Neural Networks》
目录 0 摘要/Abstract 1 简介/Introduction 2 相关工作/Related Work 3 重新思考ANN-to-SNN的转换/Rethinking ANN-to-SNN Con ...
【论文导读】- STFL: A Spatial-Temporal Federated Learning Framework for Graph Neural Networks
文章目录论文信息摘要 Contributions Methodology Graph Generation Graph Neural Network 联邦学习 Experiment 数据集 Nod ...
Machine Learning week 5 quiz: Neural Networks: Learning
Neural Networks: Learning 5 试题 1. You are training a three layer neural network and would like to us ...
《Discriminative Unsupervised Feature Learning with Exemplar Convolutional Neural Networks》阅读笔记
<Context Encoders: Feature Learning by Inpainting>阅读笔记摘要我们提出了一种基于上下文的像素预测驱动的无监督视觉特征学习算法.类似于自 ...
ICLR21： EGNN Dirichlet Energy Constrained Learning for Deep Graph Neural Networks
1. Dirichlet Energy Constrained Learning 1.1 dirichlet energy 归一化拉普拉斯矩阵+ 归一化邻接矩阵通过拉普拉斯矩阵和特征矩阵的形式求得 ...
Machine Learning week 4 quiz: Neural Networks: Representation
Neural Networks: Representation 5 试题 1. Which of the following statements are true? Check all that a ...
论文阅读——CCN GAC WORKSHOP:ISSUES WITH LEARNING IN BIOLOGICAL RECURRENT NEURAL NETWORKS
一.介绍生物中介擅长于涉及到长期短暂从属的学习任务.在学习这样的任务中的关键挑战之一是短暂信誉分配的问题:可靠地分配重要性给过去神经元状态.解决这个问题在于学习短期和长期的联合. 神经网络由周期神经 ...

A remark on the error-backpropagation learning algorithm for spiking neural networks

关于尖峰神经网络的误差反向传播学习算法的一点评论✩

摘要

脉冲感知器

A remark on the error-backpropagation learning algorithm for spiking neural networks相关推荐

最新文章

热门文章