Here we provide tutorials that demonstrate usage of JGNN in domain-specific tasks. We do not delve into inner workings of the library, as these are mainly covered in the Guidebook. Click on titles to move to corresponding documents.
This material is still a work in progress.
This is a simple example of a multilayer perceptron in action. It is a good
starting point if you want to use JGNN for traditional machine learning stuff.
Some simulated trajectory measurements are obtained across time steps
that are organized into line graphs. There are two types of trajectories:
those with full data, and those for which parts of trajectory data are missing.
The goal is to differentiate between the two types.