Using neural networks on graphs is not an easy task and there are many difficulties in using them in real-world scenarios. The first thing to think about is how NNs can ingest a graph as input. In a second step, we will see how the message passing framework can help us and we’ll apply it to homogeneous and heterogeneous graphs. Eventually, we will show GNN applied to giant graphs pointing out all the difficulties.

Speakers
Filippo Minutella - AI Engineer @ LARUS
Valerio Piccioni - AI Engineer @ LARUS

Meetup della community GraphRM