SENSE-Lab Paper Discussion Group: Meeting 3

Meeting 3

The paper presented was Johnson, M., Duvenaud, D.K., Wiltschko, A., Adams, R.P. and Datta, S.R., 2016. Composing graphical models with neural networks for structured representations and fast inference. In Advances in neural information processing systems (pp. 2946-2954).

Get it here. It can be read below as well.

If the latest, math intensive version of the paper is preferred, get it here.

Helpful Tutorials

If unfamiliar with following topics that form the basis of the paper, please look through some of the following content before the meeting which will help understand better during the meeting. The tutorials are listed in decreasing order of importance.

1. Variational Inference (mini-series)
2. Variational Autoencoders
3. Conditional Random Fields

Presentation/Notes from the meeting