Thanks for the comment.

It is a nice example that you mentioned, but I think in this article, a deterministic relation implicitly means that there is not any other variables, e.g. X4. If there exists another unknown random variable such as X4, then you can consider the setting as a probabilistic setting: i.e. given a distribution over X4, you can find the joint distribution over the others.

CS PhD student in the Laboratory of Computational Neuroscience at EPFL || Personal website: https://sites.google.com/view/modirsha

CS PhD student in the Laboratory of Computational Neuroscience at EPFL || Personal website: https://sites.google.com/view/modirsha