Graph Neural Networks: GCN w/ pure KERAS coding

You want to code a CONVOLUTION Layer for a GNN from scratch? With TensorFlow KERAS in a Jupyter NB and train your GCN to perform NODE PREDICTION?? Welcome!!

After a) GRID DATA (Vision) and b) SEQUENCE DATA (NLP - Natural Language Processing) we now switch to more complex topological data: c) GRAPH DATA!

** Great interactive explanatory of Graph and GNN basics:
https://distill.pub/2021/gnn-intro/
by Google research **

Official Google Colab NB:
https://colab.research.google.com/github/keras-team/keras-io/blob/master/examples/graph/ipynb/gnn_citations.ipynb

*** Highly recommended course for Geometric Deep Learning :
CS224W: Machine Learning with Graphs | 2021
https://www.youtube.com/watch?v=JAB_plj2rbA
by Stanford online ***
00:00 Data Topology GNN
02:30 Node Level Task
06:05 Definition GNN
08:50 Prediction Task w/ GNN
11:00 PyTorch geometric
11:20 Deep Graph Library
11:55 KERAS Jupyter NB
17:33 GNN Node Classifier

#KERAS
#GNN
#ConvolutionLayerGNN

Видео Graph Neural Networks: GCN w/ pure KERAS coding канала code_your_own_AI