Targeted Causal Elicitation
Abstract
We look at the problem of learning causal structure for a fixed downstream causal effect optimization task. In contrast to previous work which often focuses on running interventional experiments, we consider an often overlooked source of information - the domain expert. In the Bayesian setting, this amounts to augmenting the likelihood with a user model whose parameters account for possible biases of the expert. Such a model can allow for active elicitation in a manner that is most informative to the optimization task at hand.
Cite
Text
Ibrahim et al. "Targeted Causal Elicitation." NeurIPS 2022 Workshops: CML4Impact, 2022.Markdown
[Ibrahim et al. "Targeted Causal Elicitation." NeurIPS 2022 Workshops: CML4Impact, 2022.](https://mlanthology.org/neuripsw/2022/ibrahim2022neuripsw-targeted/)BibTeX
@inproceedings{ibrahim2022neuripsw-targeted,
title = {{Targeted Causal Elicitation}},
author = {Ibrahim, Nazaal and John, S. T. and Guo, Zhigao and Kaski, Samuel},
booktitle = {NeurIPS 2022 Workshops: CML4Impact},
year = {2022},
url = {https://mlanthology.org/neuripsw/2022/ibrahim2022neuripsw-targeted/}
}