Over the past decade, there has been an explosion of interest in the development and application of Approximate Bayesian Inference (ABI) methods across the statistical and machine learning (ML) communities. The rapid pace of progress in ABI has made it increasingly difficult to track development in the field, often leading to unnecessary duplication of methods and results. Without a concerted effort, statistical and ML researchers working on ABI risk drifting apart, to the detriment of both communities.
A key goal of this workshop is to strengthen connections between the statistical and ML communities engaged in ABI research by highlighting and building upon the shared foundational aim of all ABI methods: performing inference on unknown quantities given prior beliefs. By leveraging this common purpose, the workshop seeks to foster collaboration, build concrete links between the two communities, and develop novel methods to address the pressing empirical challenges faced across a wide range of scientific domains.
The workshop is organized around three recurring themes central to ABI:
- Bayesian methods for deep learning, ML, and AI
- Predictive and post-predictive Bayesian methods
- Computational approaches for Bayesian inference in complex and high-dimensional settings.
Participation is by invitation only and invited participants have been sent email invitations by ICMS.
