51 min
Seminar, Hosted by One World Approximate Bayesian Inference (OWABI) Seminar
Free-Form Flows for Physics-Informed Generative Modeling
University of Heidelberg
Recording
Abstract
This talk first introduces a useful categorisation of the sometimes confusing "generative model zoo" in terms of different change-of-variables formulas. It then shows how normalizing flows, a major architecture for generative neural networks, can be used for simulation-based Bayesian inference in the sciences. Finally, it proposes free-form flows, which simplify the incorporation of physical prior knowledge, such as rotation and translation invariance or the restriction of a distribution to a manifold, into generative models.
Topics
- Normalizing Flows
- generative modeling
- simulation-based Bayesian inference
- Physics-Informed Generative Modeling, broader topic: generative modeling
- change-of-variables formulas, search World Wide
- physical prior knowledge, broader topic: prior knowledge
- translation invariance
- manifold, search World Wide
- generative neural networks