Hacking session for ML_forecast notebook
Rudimentary notebook copying code from weatherbench
: https://github.com/pangeo-data/WeatherBench
Notebook: https://renkulab.io/gitlab/aaron.spring/s2s-ai-competition-bootstrap/-/blob/master/notebooks/ML_forecast.ipynb In its current form, the notebook is probably not useful. We should only provide such an example if reasonable, working and not misleading.
What we need is a hacking session to ensure that the ML model is trained and verified with useful data and actually improves the forecasts.
Needs input from AI and/or S2S person. Maybe Stephan Rasp.