back to list

Project: Probabilistic Circuits versus Bayesian networks

Description

This project aims to compare two different types of generative models: tractable probabilistic circuits and Bayesian networks of bounded tree-width, and potentially have tools to translate between them (when possible). Probabilistic circuits have been recently applied to a number of tasks, but there is a huge gap in the literature on how to learn their structure. One may learn certain types of Bayesian networks and then translate them to tractable circuits as means to obtain their structure.

References:

https://arxiv.org/abs/1605.03392

http://starai.cs.ucla.edu/papers/LecNoAAAI20.pdf

http://starai.cs.ucla.edu/papers/ProbCirc20.pdf

https://www.youtube.com/watch?v=2RAG5-L9R70

Details

Supervisor

Cassio de Campos

Interested?

Get in contact