Discrete (mixed) density estimators

sbi-dev / sbi

Simulation-based inference toolkit

https://sbi-dev.github.io/sbi/

Apache License 2.0

585 stars 150 forks source link

Discrete (mixed) density estimators #907

Open michaeldeistler opened 9 months ago

michaeldeistler commented 9 months ago

For discrete (mixed) parameters.

Can probably recycle a lot of code from MNLE.

janfb commented 9 months ago

Yes, MNLE implements a mixed estimator with a categorical distribution. It would be nice to extend this to discrete flows as well.

gmoss13 commented 8 months ago

Related to #968

coschroeder commented 7 months ago

happy to work on this during the hackathon

janfb commented 2 months ago

The MNLE classes are now refactored to match the API of the other build functions, including z-scoring and embedding nets. Thus, in principle, one can now also use the MNLE setup for posterior inference. However, it allows only for a single discrete column in x (theta).

An autoregressive density estimator on top would be the solution, see #1112