Abstract

This is a short course on Variational Bayes. The course will be broken up into two parts, the first corresponding to Mean Field Variational Bayes and the second on Fixed Form Variational Bayes. In the first part we will derive the key results behind Mean Field Variational Bayes and will go through in detail how to derive associated algorithms for specific models (bring pens and paper!). The second part will primarily be concerned with Gaussian variational approximations and related families of approximations. Some of the latest methods for improving variational Bayes will also be discussed time permitting.

Course Fee: Free of charge for registered attendees
Registration (Required): https://cloud.itsc.cuhk.edu.hk/webform/view.php?id=13688784
Registration Deadline: 19 May 2024