Causal Feature Learning for Utility-Maximizing Agents

David Kinney, David Watson
Proceedings of the 10th International Conference on Probabilistic Graphical Models, PMLR 138:257-268, 2020.

Abstract

Discovering high-level causal relations from low-level data is an important and challenging problem that comes up frequently in the natural and social sciences. In a series of papers, Chalupka et al. (2015, 2016a, 2016b, 2017) develop a procedure for \textit{causal feature learning} (CFL) in an effort to automate this task. We argue that CFL does not recommend coarsening in cases where pragmatic considerations rule in favor of it, and recommends coarsening in cases where pragmatic considerations rule against it. We propose a new technique, \textit{pragmatic causal feature learning} (PCFL), which extends the original CFL algorithm in useful and intuitive ways. We show that PCFL has the same attractive measure-theoretic properties as the original CFL algorithm. We compare the performance of both methods through theoretical analysis and experiments.

Cite this Paper

BibTeX
@InProceedings{pmlr-v138-kinney20a, title = {Causal Feature Learning for Utility-Maximizing Agents}, author = {Kinney, David and Watson, David}, booktitle = {Proceedings of the 10th International Conference on Probabilistic Graphical Models}, pages = {257--268}, year = {2020}, editor = {Jaeger, Manfred and Nielsen, Thomas Dyhre}, volume = {138}, series = {Proceedings of Machine Learning Research}, month = {23--25 Sep}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v138/kinney20a/kinney20a.pdf}, url = {https://proceedings.mlr.press/v138/kinney20a.html}, abstract = {Discovering high-level causal relations from low-level data is an important and challenging problem that comes up frequently in the natural and social sciences. In a series of papers, Chalupka et al. (2015, 2016a, 2016b, 2017) develop a procedure for \textit{causal feature learning} (CFL) in an effort to automate this task. We argue that CFL does not recommend coarsening in cases where pragmatic considerations rule in favor of it, and recommends coarsening in cases where pragmatic considerations rule against it. We propose a new technique, \textit{pragmatic causal feature learning} (PCFL), which extends the original CFL algorithm in useful and intuitive ways. We show that PCFL has the same attractive measure-theoretic properties as the original CFL algorithm. We compare the performance of both methods through theoretical analysis and experiments.} }
Endnote
%0 Conference Paper %T Causal Feature Learning for Utility-Maximizing Agents %A David Kinney %A David Watson %B Proceedings of the 10th International Conference on Probabilistic Graphical Models %C Proceedings of Machine Learning Research %D 2020 %E Manfred Jaeger %E Thomas Dyhre Nielsen %F pmlr-v138-kinney20a %I PMLR %P 257--268 %U https://proceedings.mlr.press/v138/kinney20a.html %V 138 %X Discovering high-level causal relations from low-level data is an important and challenging problem that comes up frequently in the natural and social sciences. In a series of papers, Chalupka et al. (2015, 2016a, 2016b, 2017) develop a procedure for \textit{causal feature learning} (CFL) in an effort to automate this task. We argue that CFL does not recommend coarsening in cases where pragmatic considerations rule in favor of it, and recommends coarsening in cases where pragmatic considerations rule against it. We propose a new technique, \textit{pragmatic causal feature learning} (PCFL), which extends the original CFL algorithm in useful and intuitive ways. We show that PCFL has the same attractive measure-theoretic properties as the original CFL algorithm. We compare the performance of both methods through theoretical analysis and experiments.
APA
Kinney, D. & Watson, D.. (2020). Causal Feature Learning for Utility-Maximizing Agents. Proceedings of the 10th International Conference on Probabilistic Graphical Models, in Proceedings of Machine Learning Research 138:257-268 Available from https://proceedings.mlr.press/v138/kinney20a.html.

Related Material