This group is interested in techniques for approximate inference in complex probabilistic models.

Schedule

Current theme is objectives for deep generative models. We meet on Fridays 11am.

Done

Variational inference

Gradient estimation, baselines, and control variates

Discrete variables and relaxations

Normalising flows

Implicit models

Bayesian deep learning

Review of variational inference (see support material at the end of the page):

Applications

Pool

Gradient Estimation

Support Material

Theses:

Tutorials and articles that read almost like tutorials:

Courses:

Gradient estimation:

Blogs: