neural
Generalised f-Mean Aggregation for Graph Neural Networks
目录概符号说明GenAgg代码 Kortvelesy R., Morad S. and Prorok A. Generalised f-mean aggregation for graph neural networks. NIPS, 2023. 概 基于 MPNN 架构的 GNN 主要在于 agg ......
Graph Condensation for Graph Neural Networks
目录概符号说明MotivationGCOND代码 Jin W., Zhao L., Zhang S., Liu Y., Tang J. and Shah N. Graph condensation for graph neural networks. ICLR, 2022. 概 图上做压缩的工作. ......
BIgdataAIML-IBM-A neural networks deep dive - An introduction to neural networks and their programming
https://developer.ibm.com/articles/cc-cognitive-neural-networks-deep-dive/ By M. Tim Jones, Published July 23, 2017 Neural networks have been around f ......
Non-local Neural Networks 第一次将自注意力用于cv
Non-local Neural Networks * Authors: [[Xiaolong Wang]], [[Ross Girshick]], [[Abhinav Gupta]], [[Kaiming He]] Local library 初读印象 comment:: (NonLocal)过去 ......
Real-Time Single Image and Video Super-Resolution Using an Efficient Sub-Pixel Convolutional Neural Network
Real-Time Single Image and Video Super-Resolution Using an Efficient Sub-Pixel Convolutional Neural Network * Authors: [[Wenzhe Shi]], [[Jose Caballer ......
Asymmetric Non-Local Neural Networks for Semantic Segmentation 非对称注意力
Asymmetric Non-Local Neural Networks for Semantic Segmentation * Authors: [[Zhen Zhu]], [[Mengdu Xu]], [[Song Bai]], [[Tengteng Huang]], [[Xiang Bai]] ......
SiReN Sign-Aware Recommendation Using Graph Neural Networks论文阅读笔记
Abstract 目前使用GNN的推荐系统主要利用高评分的正向用户-物品交互信息。但是如何利用低评分来表示用户的偏好是一个挑战，因为低评分仍然可以提供有用的信息。所以在本文中提出了基于GNN模型的有符号感知推荐系统SiReN，SiReN有三个关键组件 构造一个符号二部图更精确的表示用户的偏好，分为两 ......
PANE-GNN Unifying Positive and Negative Edges in Graph Neural Networks for Recommendation论文阅读笔记
Abstract 目前利用GNN的推荐系统主要关注用户的正面反馈，而忽略了负面反馈提供的见解。于是我们提出了PANG- GNN，该模型将图神经网络的正面和负面边统一在一起。PANG-GNN首先将原始评分图根据正面和负面反馈划分为两个不同的二分图。接下来分别使用两个独立的嵌入，即感兴趣嵌入和无兴趣嵌入 ......
A fast and simple algorithm for training neural probabilistic language models
目录概Noise contrastive estimation Mnih A. and Teh Y. W. A fast and simple algorithm for training neural probabilistic language models. ICML, 2012. 概 NCE ......
Predicting gene expression from histone modifications with self-attention based neural networks and transfer learning
Predicting gene expression from histone modifications with self-attention based neural networks and transfer learning Yuchi Chen 1, Minzhu Xie 1, Jie ......
神经网络入门篇：详解搭建神经网络块（Building blocks of deep neural networks）
搭建神经网络块 这是一个层数较少的神经网络，选择其中一层（方框部分），从这一层的计算着手。在第\(l\)层有参数\(W^{[l]}\)和\(b^{[l]}\)，正向传播里有输入的激活函数，输入是前一层\(a^{[l-1]}\)，输出是\(a^{[l]}\)，之前讲过\(z^{[l]} =W^{[l] ......
Guo_AD-NeRF_Audio_Driven_Neural_Radiance_Fields_for_Talking_Head_Synthesis_ICCV_2021_paper
可以看看这个向量场的虚拟人像的效果. 看论文第三章: 3.2: F_theta是一个神经网络, a是声音 d 是view direction, x是3d location. 普通的向量场是 F_theta: d,x > (c,σ) 表示d是一个方向, 表示观看者水平的偏移角度和数值的偏移角度. x是 ......
Graph Neural Networks with Learnable and Optimal Polynomial Bases
目录概符号说明MotivationFavardGNN代码 Guo Y. and Wei Z. Graph neural networks with learnable and optimal polynomial bases. ICML, 2023. 概 自动学多项式基的谱图神经网络. 符号说明 \ ......
[论文速览] R-Drop@ Regularized Dropout for Neural Networks
Pre title: R-Drop: Regularized Dropout for Neural Networks accepted: NeurIPS 2021 paper: https://arxiv.org/abs/2106.14448 code: https://github.com/dro ......
论文：Predicting the performance of green stormwater infrastructure using multivariate long short-term memory (LSTM) neural network
题目“Predicting the performance of green stormwater infrastructure using multivariate long short-term memory (LSTM) neural network” (Al Mehedi 等, 2023, ......
The Hello World of Deep Learning with Neural Networks
The Hello World of Deep Learning with Neural Networks dlaicourse/Course 1 - Part 2 - Lesson 2 - Notebook.ipynb at master · lmoroney/dlaicourse (github ......
The Hello World of Deep Learning with Neural Networks
The Hello World of Deep Learning with Neural Networks dlaicourse/Course 1 - Part 2 - Lesson 2 - Notebook.ipynb at master · lmoroney/dlaicourse (github ......
CrossEntropyLoss: RuntimeError: expected scalar type Float but found Long neural network
错误分析 这个错误通常指的是期望接受的参数类型是Float， 但是程序员传入的是Int 。 通常会需要我们去检查传入的 input 和 target 的数据类型有没有匹配。在传入的数据中，通常 input 希望是 Float 类型，target 是 Int 类型。 但是通常也许会发现传入的参数是符合 ......
Graph Neural Networks with Diverse Spectral Filtering
目录概符号说明DSF代码 Guo J., Huang K, Yi X. and Zhang R. Graph neural networks with diverse spectral filtering. WWW, 2023. 概 为每个结点赋予不同的多项式系数. 符号说明 \(\mathcal{ ......
Convolutional Neural Networks on Graphs with Chebyshev Approximation, Revisited
目录概符号说明MotivationChebNetII代码 He M., Wei Z. and Wen J. Convolutional neural networks on graphs with chebyshev approximation, revisited. NIPS, 2022. 概 作 ......
How Powerful are Spectral Graph Neural Networks?
目录概符号说明Spectral GNNChoice of Basis for Polynomial FiltersJacobiConv代码 Wang X. and Zhang M. How powerful are spectral graph neural networks? ICML, 2022 ......
神经网络入门篇：神经网络的梯度下降（Gradient descent for neural networks）
神经网络的梯度下降 在这篇博客中，讲的是实现反向传播或者说梯度下降算法的方程组 单隐层神经网络会有\(W^{[1]}\)，\(b^{[1]}\)，\(W^{[2]}\)，\(b^{[2]}\)这些参数，还有个\(n_x\)表示输入特征的个数，\(n^{[1]}\)表示隐藏单元个数，\(n^{[2]} ......
Decoupling the Depth and Scope of Graph Neural Networks
目录概符号说明Shadow-GNN代码 Zeng H., Zhang M., Xia Y., Srivastava A., Malevich A., Kannan R., Prasanna V., Jin L. and Chen R. Decoupling the depth and scope o ......
SuperGlue: Learning Feature Matching with Graph Neural Networks论文笔记
SuperGlue: Learning Feature Matching with Graph Neural Networks 源码: github.com/magicleap/SuperGluePretrainedNetwork 背景： 主要解决图像中点之间的对应关系。 主要方法： 上图为该方法的 ......
Neural Networks投稿要求总结
自用，NN投稿要求，相关的部分的中文版翻译，原文链接：https://www.sciencedirect.com/journal/neural-networks/publish/guide-for-authors Neural Networks 投稿要求 介绍 国际神经网络学会、欧洲神经网络学会和日 ......
神经网络入门篇：详解计算一个神经网络的输出（Computing a Neural Network's output）
一个神经网络的输出 首先，回顾下只有一个隐藏层的简单两层神经网络结构： 图1.3.1 其中，\(x\)表示输入特征，\(a\)表示每个神经元的输出，\(W\)表示特征的权重，上标表示神经网络的层数（隐藏层为1），下标表示该层的第几个神经元。这是神经网络的符号惯例，下同。 神经网络的计算 关于神经网络 ......
Graph Neural Networks with Adaptive Residual
目录概符号说明AirGNN代码 Liu X., Ding J., Jin W., Xu H., Ma Y., Liu Z. and Tang J. Graph neural networks with adaptive residual. NIPS, 2021. 概 基于 UGNN 框架的一个更加鲁 ......
Is Homophily a Necessity for Graph Neural Networks?
目录概 Ma Y., Liu X., Shah N. and Tang J. Is homophily a necessity for graph neural networks? ICLR, 2022. 概 探究 Homophily 假设 (即相互连接的结点相似) 对于 GCN 发挥效果是否是必须 ......
华为最高学术成果发表 —— 《Nature》正刊发表论文《Accurate medium-range global weather forecasting with 3D neural networks》
论文《Accurate medium-range global weather forecasting with 3D neural networks》的《Nature》地址： https://www.nature.com/articles/s41586-023-06185-3.pdf 论文的代码地 ......
GraphPrompt: Unifying Pre-Training and Downstream Tasks for Graph Neural Networks
目录概符号说明GraphPrompt代码 Liu Z., Yu X., Fang Y. and Zhang X. GraphPrompt: Unifying pre-training and downstream tasks for graph neural networks. WWW, 2023. ......