Graph Convolutional Network (GCN) Paper Explained

In this video we do a paper walk through of the GCN (Graph Convolutional Network) paper that is the most cited and one of the impactful in graph neural networks. We understand how they derived it and how it works in detail.

Big thanks to MakinaRocks for sponsoring this video, and I encourage you to check out Link which is a jupter lab extension they have developed to make notebooks more intuitive and easier to use. A few of the features I’ve found useful are:
1️⃣ Caching result from running cells in notebooks to avoid re-running them
2️⃣ Visualizing cell dependencies
3️⃣ Integrated version control

Check it out⬇️
https://bit.ly/3F4COvv

Link Demo Video 👉 https://youtu.be/uM2uPG-1eQQ
Link Documentation 👉https://makinarocks.gitbook.io/link/

Timestamps:
0:00 - Introduction
1:00 - Sponsored Segment: Link
2:09 - Abstract and overview
4:13 - Introduction to the problem
8:53 - How GCNs work
13:47 - Theory of GCN derivation
21:25 - GCN Node classification example
24:33 - Conclusions & Results
25:13 - Ending thoughts

#link #gcn