Semi-symbolic inference for efficient streaming probabilistic programming
Abstract
A streaming probabilistic program receives a stream of observations and produces a stream of distributions that are conditioned on these observations. Efficient inference is often possible in a streaming context using Rao-Blackwellized particle filters (RBPFs), which exactly solve inference problems when possible and fall back on sampling approximations when necessary. While RBPFs can be implemented by hand to provide efficient inference, the goal of streaming probabilistic programming is to automatically generate such efficient inference implementations given input probabilistic programs.
Document Details
- Document Type
- Pub Defense Publication
- Publication Date
- Oct 31, 2022
- Source ID
- 10.1145/3563347
Entities
People
- Charles Yuan
- Eric Atkinson
- Guillaume Baudart
- Louis Mandel
- Michael Carbin
Organizations
- IBM Research
- Massachusetts Institute of Technology
- Office of Naval Research
- PSL Research University