Deep Learning for Dynamic Graph

Introduction.

It is well understood that adding the time dimension to each and every component of the graph helps us in converting the traditional graph into a dynamic graph.?This time dimension makes the dynamic graph capable to understand the complex behavior of real-world entities (like human beings in a social network, financial activities, stock-market-related activities, gene coding protein in a PPI network, etc.).

After the development of deep-learning-based systems, people also started their efforts toward the development of efficient deep neural network architectures for dynamic graphs. In this process, people have focused on the following set of problems.

1. How to effectively capture the temporal events for the deep learning system?
2. How to select between discrete and continuous processes to properly justify the goal and use of the deep learning system?
3. How to apply the deep-learning for temporal components?
4. How to apply the attention mechanism in such cases?

I have compiled the following set of tutorials to properly explain the use of deep learning for dynamic graph.

1.Dynamic Graph Neural Networks Part-1

Contains - [ ?Transition from Static to Dynamic Graph Neural Networks; Dynamic Graph Neural Network; Temporal Graph Networks; Deep Learning for Dynamic Temporal Graph Networks]

2.Dynamic Graph Neural Networks Part-2

Contains - [Deep Learning for Dynamic Temporal Graph Networks; Temporal Graph Embeddings; Temporal Graph Attention.]

3.?Graph Attention Networks; Multi-Head Attention

Contains - [Graph Attention Networks; Multi-Head Attention in Graph Networks.]

4.?Static Vs Dynamic Graph Neural Networks (For those who wants additional basic information about Dynamic Graph)

Contains - [Deep Learning using Static Graph; Graph Neural Network (GNN) Vs Graph Convolutional Neural Networks (GCN); Intro to Dynamic Graph?]

References.

1. Rossi, Emanuele, Ben Chamberlain, Fabrizio Frasca, Davide Eynard, Federico Monti, and Michael Bronstein. "Temporal graph networks for deep learning on dynamic graphs 2020."?arXiv preprint arXiv:2006.10637?(2006).
2. Velickovic, Petar, Guillem Cucurull, Arantxa Casanova, Adriana Romero, Pietro Lio, and Yoshua Bengio. "Graph attention networks."?stat?1050 (2017): 20.
3. Vaswani, Ashish, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N. Gomez, ?ukasz Kaiser, and Illia Polosukhin. "Attention is all you need."?Advances in neural information processing systems?30 (2017).