Here are collected a bunch of notes, to clarify investigation of trivial and non-trivial aspects.
Non-trivial results are collected to be reused. Trivial ones to develop confidence (and not to worry again about)
extra:
- notes on further options that will be available after we'll have a baseline ready
- notes on variational Bayes
- I don't consider a promising approach, but better to keep in mind AC