AI as Statistical Methods for Imperfect Theories

Abstract

Science has progressed by reasoning on what models could not predict because they were missing important ingredients. And yet without correct models, standard statistical methods for scientific evidence are not sound. Here I argue that machine-learning methodology provides solutions to ground reasoning about empirically evidence more on models’ predictions, and less on their ingredients.

PDF NeurIPSW OpenReview Semantic Scholar

Cite

Text

Varoquaux. "AI as Statistical Methods for Imperfect Theories." NeurIPS 2021 Workshops: AI4Science, 2021.

Markdown

[Varoquaux. "AI as Statistical Methods for Imperfect Theories." NeurIPS 2021 Workshops: AI4Science, 2021.](https://mlanthology.org/neuripsw/2021/varoquaux2021neuripsw-ai/)

BibTeX

@inproceedings{varoquaux2021neuripsw-ai,
  title     = {{AI as Statistical Methods for Imperfect Theories}},
  author    = {Varoquaux, Gael},
  booktitle = {NeurIPS 2021 Workshops: AI4Science},
  year      = {2021},
  url       = {https://mlanthology.org/neuripsw/2021/varoquaux2021neuripsw-ai/}
}