A Tutorial on Parametric Variational Inference
It provides an accessible introduction to parametric variational inference for researchers and practitioners in machine learning and statistics, focusing on recent developments rather than being incremental.
This tutorial introduces parametric variational inference, a method for approximating Bayesian posteriors using optimization, which has become the preferred approach for high-dimensional models and large datasets due to recent scalability advances.
Variational inference uses optimization, rather than integration, to approximate the marginal likelihood, and thereby the posterior, in a Bayesian model. Thanks to advances in computational scalability made in the last decade, variational inference is now the preferred choice for many high-dimensional models and large datasets. This tutorial introduces variational inference from the parametric perspective that dominates these recent developments, in contrast to the mean-field perspective commonly found in other introductory texts.