Skip to content

Latest commit

 

History

History
37 lines (27 loc) · 1.21 KB

README.md

File metadata and controls

37 lines (27 loc) · 1.21 KB

MAD Mix

This is a python package that implements MAD Mix, a flow-based variational methodology to learn discrete distributions.

Directory roadmap

Each subdirectory contains README files. Generally:

  • src/ contains the source code for MAD Mix, including instantiations for discrete-only, continuous-only (as in Xu et al. (2022)), and mixed discrete and continuous models
  • examples/ has examples where MAD Mix and other methods are used to learn different distributions, some purely discrete and some including continuous variables as well.

Citing MAD Mix

If you find our code useful, consider citing our paper.

BibTeX code for citing MAD Mix

@inproceedings{diluvi2024madmix,
  title={Mixed variational flows for discrete variables},
  author={{Gian Carlo} Diluvi and Benjamin Bloem-Reddy and Trevor Campbell},
  booktitle={International Conference on Artificial Intelligence and Statistics},
  year={2024}
  }

APA

Diluvi, G.C., Bloem-Reddy, B., and Campbell, T. Mixed variational flows for discrete variables. In International Conference on Artificial Intelligence and Statistics, 2024.