As soon as I saw the title, my reaction was "YESSSS!"... A while ago, I was looking for something like this, but I soon realized that deep learning on graphs is still comparatively new, and not many python modules exist for it.
I'd love to tinker with this. What is a classic problem or demo for graph neural networks? Is there an equivalent to the MNIST problem for GNNs that would make a good learning exercise?
How is the data fed into the model? Are the graphs represented by a dictionary or an adjacency matrix (or neither)? Does anyone have more links I can read more about graph based deep learning?
As soon as I saw the title, my reaction was "YESSSS!"... A while ago, I was looking for something like this, but I soon realized that deep learning on graphs is still comparatively new, and not many python modules exist for it.
Glad you like it. Now go and build crazy stuff!
I'd love to tinker with this. What is a classic problem or demo for graph neural networks? Is there an equivalent to the MNIST problem for GNNs that would make a good learning exercise?
How is the data fed into the model? Are the graphs represented by a dictionary or an adjacency matrix (or neither)? Does anyone have more links I can read more about graph based deep learning?
The official docs have some of the details: https://danielegrattarola.github.io/spektral/data/
More about Graph NNs: https://arxiv.org/pdf/1812.08434.pdf
Thank you!!
Machine Learning in Complex Networks
https://link.springer.com/book/10.1007/978-3-319-17290-3
Front Matter - https://link.springer.com/content/pdf/bfm%3A978-3-319-17290-...
Vectorize the adjacency matrix.
YEEES I will finally be able to complete a college assignment that I postponed