Deep Learning - 文章分类 - veager

摘要：GloVe 阅读全文

posted @ 2023-02-23 17:40 veager 阅读(13) 评论(0) 推荐(0)

摘要：1 Temporal Convolutional Network (TCN) 时间卷积网络(TCN)：结构+pytorch代码, site 阅读全文

posted @ 2022-08-29 19:25 veager 阅读(43) 评论(0) 推荐(0)

摘要：Implement Graph Neural Networks Model 1 Graph Convolutional Networks (GCN) 1.1 Formulation The propagation function $f$: $$ \mathbf{H}^{(l+1)} = f \le 阅读全文

posted @ 2022-07-25 20:49 veager 阅读(64) 评论(0) 推荐(0)

网络优化与正则化

摘要：1. 网络优化 1.1 网络结构多样性 1.2 高维变量的非凸优化鞍点（Saddle Point）局部最小值（Local Minima）驻点（Stationary Point）在高维空间中大部分驻点都是鞍点平坦最小值（Flat Minima）尖锐最小值（Sharp Minima）局部最阅读全文

posted @ 2022-07-22 10:48 veager 阅读(463) 评论(0) 推荐(0)

Machine Learning with Graphs : 9 Theory of Graph Neural Networks

摘要：Stanford CS224W : Machine Learning with Graphs, Fall 2021: 9 Theory of Graph Neural Networks site 1. Setting-up GNN Prediction Tasks 1.1 GNN Training 阅读全文

posted @ 2022-07-17 19:41 veager 阅读(472) 评论(0) 推荐(0)

Machine Learning with Graphs : 8 Applications of Graph Neural Networks

摘要：Stanford CS224W : Machine Learning with Graphs, Fall 2021: 8 Applications of Graph Neural Networks site 1. GNN Augmentation and Training 2. Stacking G 阅读全文

posted @ 2022-07-17 19:39 veager 阅读(93) 评论(0) 推荐(0)

Machine Learning with Graphs : 7 Graph Neural Networks 2: Design Space

摘要：Stanford CS224W : Machine Learning with Graphs, Fall 2021: 7 Graph Neural Networks 2: Design Space site 1. A General Perspective on GNNs 1.1 A General 阅读全文

posted @ 2022-07-14 17:32 veager 阅读(111) 评论(0) 推荐(0)

Machine Learning with Graphs : 6 Graph Neural Networks

摘要：Stanford CS224W : Machine Learning with Graphs, Fall 2021: 6 Graph Neural Networks site 1. Overview 1.1 Node Embeddings Limitations of shallow embeddi 阅读全文

posted @ 2022-07-13 20:53 veager 阅读(64) 评论(0) 推荐(0)

Machine Learning with Graphs : 5 Message Passing and Node Classification

摘要：Stanford CS224W : Machine Learning with Graphs, Fall 2021: 5 Message Passing and Node Classification site 1. Overview 2. Leverage Node Correlations in 阅读全文

posted @ 2022-07-13 17:39 veager 阅读(23) 评论(0) 推荐(0)

Machine Learning with Graphs : 4 Graph as Matrix: PageRank, Random Walks and Embeddings

摘要：Stanford CS224W : Machine Learning with Graphs, Fall 2021: 4 Graph as Matrix: PageRank, Random Walks and Embeddings site 1. Overview Treating a graph 阅读全文

posted @ 2022-07-13 16:09 veager 阅读(89) 评论(0) 推荐(0)

Machine Learning with Graphs : 3 Node Embeddings

摘要：Machine Learning with Graphs : 3 Node Embeddings 1. Overview Graph Representation Learning alleviates the need to do feature engineering every single 阅读全文

posted @ 2022-07-10 19:35 veager 阅读(155) 评论(0) 推荐(0)

Graph PyTorch Geometric 库

摘要：PyTorch Geometric (PyG) 1. 安装安装：site import os import torch print(torch.__version__) os.environ['TORCH'] = torch.__version__ ! pip install -q torch-s 阅读全文

posted @ 2022-07-09 10:50 veager 阅读(546) 评论(0) 推荐(0)

Machine Learning with Graphs : 1. Introduction

摘要：Stanford CS224W : Machine Learning with Graphs, Fall 2021: 1 Introduction; Machine Learning for Graphs site 1. Classic Graph ML tasks Node classificat 阅读全文

posted @ 2022-07-08 20:05 veager 阅读(69) 评论(0) 推荐(0)

PyTorch 正则化，读写模型

摘要：PyTorch 正则化，读写模型 1. 正则化 1.1 权重衰减 Weight Decay 权重衰减（Weight Decay）在每次参数更新时，引入一个衰减系数。 \[ \theta_{t} \leftarrow(1-\beta) \theta_{t-1}-\alpha g_{t} \] 其中，\ 阅读全文

posted @ 2022-05-27 00:13 veager 阅读(705) 评论(0) 推荐(0)

PyTorch 损失函数

摘要：PyTorch 损失函数 Loss Functions 0 概述 PyTorch 中，损失函数有两种形式，与激活函数类似：层（模块）的形式：需要先定义，再使用。为 torch.nn 模块下的类，官方文档函数形式：直接使用。为 torch.nn.functional 模块中的函数，官方文档 PyT 阅读全文

posted @ 2022-05-26 19:23 veager 阅读(865) 评论(0) 推荐(0)

激活函数 Activation Functions

摘要：PyTorch 激活函数 Activation Functions 0. 概述 PyTorch 中，激活函数有两种形式：层（模块）的形式：需要先定义，再使用。为 torch.nn 模块下的类，官方文档函数形式：直接使用。为 torch.nn.functional 模块或 torch.nn 模块中阅读全文

posted @ 2022-05-26 19:22 veager 阅读(390) 评论(0) 推荐(0)

PyTorch Lightning 教程

摘要：PyTorch Lightning 教程 1. TorchMetrics 库指标评估：官网阅读全文

posted @ 2022-05-25 16:30 veager 阅读(24) 评论(0) 推荐(0)

veager

文章分类 - Deep Learning