课程简介:
复杂的数据可以表示为对象之间的关系图。这种网络是对社会、技术和生物系统进行建模的基本工具。本课程重点关注海量图分析所面临的计算、算法和建模挑战。通过研究底层图结构及其特征,向学生介绍机器学习技术和数据挖掘工具,这些工具有助于揭示对各种网络的见解。
主题包括:表示学习和图神经网络;万维网算法;对知识图进行推理;影响力最大化;疾病爆发检测、社交网络分析。
课程官网:
(最新秋季课程)https://web.stanford.edu/class/cs224w/
(春季课程)https://snap.stanford.edu/class/cs224w-2023/
视频资源不开放
视频:
2023年春季斯坦福CS224W《图机器学习》 ——同济子豪兄中文精讲:
-
https://github.com/TommyZihao/zihao_course/tree/main/CS224W
-
https://space.bilibili.com/1900783/channel/collectiondetail?sid=915098
-
【【中英字幕】2021年春季斯坦福大学Stanford CS224W《图机器学习Machine Learning with Graphs》课程】 https://www.bilibili.com/video/BV1s54y1H76H/?share_source=copy_web&vd_source=be2dbc6e276c34a5d143c75d66d47b6d
2023年秋季斯坦福CS224W《图机器学习》课程团队:
2023年秋季斯坦福CS224W《图机器学习》课程内容大纲:
DateDescriptionOptional ReadingsEventsDeadlinesTue, 9/261. Introduction
[slides]
Thu, 9/282. Node embeddings
[slides]
-
DeepWalk: Online Learning of Social Representations
-
node2vec: Scalable Feature Learning for Networks
-
Network Embedding as Matrix Factorization
Colab 0, Colab 1 out
Tue, 10/33. Graph neural networks
[slides]
-
Geometric Deep Learning: the Erlangen Programme of ML
-
Semi-Supervised Classification with Graph Convolutional Networks
Thu, 10/54. A general perspective on GNNs
[slides]
-
Design Space of Graph Neural Networks
-
Inductive Representation Learning on Large Graphs
-
Graph Attention Networks
Homework 1 out LaTeX template
Tue, 10/105. GNN augmentation and training
[slides]
-
Hierarchical Graph Representation Learning with Differentiable Pooling
Thu, 10/126. Theory of GNNs
[slides]
-
How Powerful Are Graph Neural Networks?
Colab 2 out
Tue, 10/177. Heterogenous graphs
[slides]
-
Modeling Relational Data with Graph Convolutional Networks
-
Heterogeneous Graph Transformer
Colab 1 dueThu, 10/198. Knowledge graphs
[slides]
-
Translating Embeddings for Modeling Multi-relational Data
-
Learning Entity and Relation Embeddings for Knowledge Graph Completion
-
Embedding Entities and Relations for Learning and Inference in Knowledge Bases
-
Complex Embeddings for Simple Link Prediction
-
RotatE: Knowledge Graph Embedding by Relational Rotation in Complex Space
Homework 2 out LaTeX templateHomework 1 dueTue, 10/249. Reasoning over knowledge graphs
[slides]
-
Embedding Logical Queries on Knowledge Graphs
-
Query2box: Reasoning over Knowledge Graphs in Vector Space Using Box Embeddings
-
Traversing Knowledge Graphs in Vector Space
Project Proposal dueThu, 10/2610. Fast neural subgraph matching
[slides]
-
Network Motifs: Simple Building Blocks of Complex Networks
-
Neural Subgraph Matching
-
SPMiner: Frequent Subgraph Mining by Walking in Order Embedding Space
Colab 3 outColab 2 dueTue, 10/3111. GNNs for recommenders
[slides]
-
Neural Graph Collaborative Filtering
-
LightGCN: Simplifying and Powering Graph Convolution Network for Recommendation
-
Graph Convolutional Neural Networks for Web-Scale Recommender Systems
Thu, 11/212. Deep generative models for graphs
[slides]
-
GraphRNN: Generating Realistic Graphs with Deep Auto-regressive Models
-
Graph Convolutional Policy Network for Goal-Directed Molecular Graph Generation
Homework 3 out LaTeX templateHomework 2 dueTue, 11/7ELECTION DAY – NO CLASS
Thu, 11/913. Advanced topics in GNNs
[slides]
-
Position-aware Graph Neural Networks
-
Identity-aware Graph Neural Networks
-
Adversarial Attacks on Neural Networks for Graph Data
Colab 4 outColab 3 due
Project Milestone dueTue, 11/1414. Graph Transformers
[slides]
Thu, 11/1615. Scaling to large graphs
[slides]
-
Cluster-GCN: An Efficient Algorithm for Training Deep and Large Graph Convolutional Networks
-
Simplifying Graph Convolutional Networks
Colab 5 outHomework 3 dueTue, 11/21BREAK
Tue, 11/23BREAK
Tue, 11/2816. Geometric deep learning
[slides]
-
SchNet: A continuous-filter convolutional neural network for modeling quantum interactions
-
Equivariant message passing for the prediction of tensorial properties and molecular spectra
-
Tensor field networks: Rotation- and translation-equivariant neural networks for 3D point clouds
-
GeoDiff: A Geometric Diffusion Model for Molecular Conformation Generation
Wed, 11/29Exam
11/29 5 PM – 12/1 5 AMThu, 11/3017. Link Prediction and Causality
[slides]
Colab 4 dueTue, 12/518. Algorithmic reasoning with GNNs
[slides]
-
LIME: Local Interpretable Model-Agnostic Explanations
-
A Unified Approach to Interpreting Model Predictions
-
GNNExplainer
-
Explainability in Graph Neural Networks: A Taxonomic Survey
-
Trustworthy Graph Neural Networks
-
GraphFramEx
Colab 5 dueThu, 12/719. Conclusion
[slides]
Thu, 12/14
Project Report due
2023年春季斯坦福CS224W《图机器学习》课程团队:
2023年春季斯坦福CS224W《图机器学习》课程内容大纲:
微信群 公众号