1 March 2021 to 9 April 2021
Institut Henri Poincaré
Europe/Paris timezone

Simulation-efficient marginal posterior estimation with swyft

Not scheduled
20m
Amphitheater Darboux (Institut Henri Poincaré)

Amphitheater Darboux

Institut Henri Poincaré

11, Rue Pierre et Marie Curie 75005 Paris
Poster

Description

We present algorithms (a) for nested neural likelihood-to-evidence ratio estimation, and (b) for simulation reuse via an inhomogeneous Poisson point process cache of parameters and corresponding simulations. Together, these algorithms enable automatic and extremely simulator efficient estimation of marginal and joint posteriors. The algorithms are applicable to a wide range of physics and astronomy problems and typically offer an order of magnitude better simulator efficiency than traditional likelihood-based sampling methods. Our approach is an example of likelihood-free inference, thus it is also applicable to simulators which do not offer a tractable likelihood function. Simulator runs are never rejected and can be automatically reused in future analysis. As functional prototype implementation we provide the open-source software package swyft. https://github.com/undark-lab/swyft

Primary author

Benjamin Miller (University of Amsterdam)

Co-authors

Dr Alex Cole (University of Amsterdam) Dr Gilles Louppe (University of Liège) Dr Christoph Weniger (University of Amsterdam)

Presentation Materials

There are no materials yet.