Feat/run in docker

[review-notebook-app[bot]]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[alronlam]

Caveat: I didn't get the time to really look at the whole docs in-depth.

But just to recap on a high-level, all of these efforts are mainly to cater for the case where a data scientist just wants to train and rollout on different countries/years without changing any of the underlying feature engg/modelling right?

If so, might want to have a short description somewhere to make this explicit just to guide external people, especially those who might not be familiar with Docker. That is, something like:

• If you're looking to simply run the training/rollout on different countries/years, follow this...
• If you want to do ML experiments and really change the underlying modelling and feature engineering code, then follow this... (regular setup or docker)

Cause I'm imagining there might be a demographic of data scientists who are not familiar with Docker, and our README now contains so many instructions. You might think that you need to do all of them.