We have prepared jupyter notebooks as supplementary material to the "Subgroup discovery, rare-phenomena challenge, and domain of applicability" lecture. 
Go to https://nomad-lab.eu/AItutorials (the general landing page of the NOMAD AI-toolkit) and look for "Identifying Domains of Applicability of Machine-Learning Models for Materials Science" (same title as the paper). 
You can walk through the content presented in the paper, can reproduce the result and can modify the input parameters to the algorithms to obtain new results.
From the same page, you can also access the notebook "NOMAD 2018 Kaggle research competition" which presents the winning models (descriptors and regressors), again with possibility to modify the input parameters."

Important links
https://analytics-toolkit.nomad-coe.eu/public/user-redirect/notebooks/tutorials/domain_of_applicability.ipynb
https://analytics-toolkit.nomad-coe.eu/public/user-redirect/notebooks/tutorials/kaggle_competition.ipynb