ML Anthology

Ad-Hoc Bayesian Program Learning

Eli Sennesh
NeurIPSW 2019
/neuripsw/2019/sennesh2019neuripsw-adhoc/

Abstract

Bayesian program learning provides a general approach to human-level concept learning in artificial intelligence. However, most priors over powerful programming languages make searching for a high-scoring program intractable, and therefore cognitively unrealistic. We hypothesize that an efficient learner searches programs which efficiently generate a likelihood by running to completion, and model this hypothesis with an ad-hoc proposal for programs. Our proposal works backwards from observations to find programs which quickly generate similar results.

PDF NeurIPSW OpenReview Semantic Scholar

Cite

Text

Sennesh. "Ad-Hoc Bayesian Program Learning." NeurIPS 2019 Workshops: Program_Transformations, 2019.

Markdown

[Sennesh. "Ad-Hoc Bayesian Program Learning." NeurIPS 2019 Workshops: Program_Transformations, 2019.](https://mlanthology.org/neuripsw/2019/sennesh2019neuripsw-adhoc/)

BibTeX

@inproceedings{sennesh2019neuripsw-adhoc,
  title     = {{Ad-Hoc Bayesian Program Learning}},
  author    = {Sennesh, Eli},
  booktitle = {NeurIPS 2019 Workshops: Program_Transformations},
  year      = {2019},
  url       = {https://mlanthology.org/neuripsw/2019/sennesh2019neuripsw-adhoc/}
}