For Graph ML we make a deep dive to code LINK Prediction on Graph Data sets with DGL and PyG. We examine the main ideas behind LINK Prediction and how to code a link prediction example in PyG and DGL - Deep Graph Library. DGL - Easy Deep Learning on Graphs with framework agnostic coding (either PyG or TensorFlow2).

A GNN-based link prediction model represents the likelihood of connectivity between two nodes u and v as a function of their node representation computed from the multi-layer GNN.

Training a link prediction model involves comparing the scores between nodes connected by an edge against the scores between an arbitrary pair of nodes. For example, given an edge connecting u and v, we encourage the score between node u and v to be higher than the score between node u and a sampled node v′ from an arbitrary noise distribution v′∼Pn(v). Such methodology is called "negative sampling".

I recommend:
https://docs.dgl.ai/en/0.9.x/guide/tr...

My GraphSAGE explanation video:
   • Unlocking the Potential of Message Pa...  

Thanks to Canva (canva.com) for providing of free version of canva to the global community.  @canva 

#datascience
#machinelearning
#graphs
#dgl