Estimating individual-level optimal causal interventions combining causal models and machine learning models

Keisuke Kiritoshi, Tomonori Izumitani, Kazuki Koyama, Tomomi Okawachi, Keisuke Asahara, Shohei Shimizu
Proceedings of The KDD'21 Workshop on Causal Discovery, PMLR 150:55-77, 2021.

Abstract

We introduce a new statistical causal inference method to estimate individual-level optimal causal intervention, that is, to which value we should set the value of a certain variable of an individual to obtain a desired value of another variable. This is defined as an optimization problem to minimize the error between a desired value and the value that would have been attained under the setting for the individual. To solve the optimization problem, we first train a machine learning model to predict the value of an objective variable and then estimate the causal structure of variables. We then combine the machine learning model and causal structure into a single causal model to estimate counterfactual value of the predicted objective variable. This is effective in achieving a more accurate estimation of individual-level optimal causal intervention. We further propose a gradient descent algorithm to compute the optimal causal intervention. Our method is generally applicable to continuous variables that are linearly and non-linearly related. In experiments, we evaluate the effectiveness of our method using artificial data generated by non-linear causal structures and real data.

Cite this Paper

BibTeX
@InProceedings{pmlr-v150-kiritoshi21a, title = {Estimating individual-level optimal causal interventions combining causal models and machine learning models}, author = {Kiritoshi, Keisuke and Izumitani, Tomonori and Koyama, Kazuki and Okawachi, Tomomi and Asahara, Keisuke and Shimizu, Shohei}, booktitle = {Proceedings of The KDD'21 Workshop on Causal Discovery}, pages = {55--77}, year = {2021}, editor = {Le, Thuc Duy and Li, Jiuyong and Cooper, Greg and Triantafyllou, Sofia and Bareinboim, Elias and Liu, Huan and Kiyavash, Negar}, volume = {150}, series = {Proceedings of Machine Learning Research}, month = {15 Aug}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v150/kiritoshi21a/kiritoshi21a.pdf}, url = {https://proceedings.mlr.press/v150/kiritoshi21a.html}, abstract = {We introduce a new statistical causal inference method to estimate individual-level optimal causal intervention, that is, to which value we should set the value of a certain variable of an individual to obtain a desired value of another variable. This is defined as an optimization problem to minimize the error between a desired value and the value that would have been attained under the setting for the individual. To solve the optimization problem, we first train a machine learning model to predict the value of an objective variable and then estimate the causal structure of variables. We then combine the machine learning model and causal structure into a single causal model to estimate counterfactual value of the predicted objective variable. This is effective in achieving a more accurate estimation of individual-level optimal causal intervention. We further propose a gradient descent algorithm to compute the optimal causal intervention. Our method is generally applicable to continuous variables that are linearly and non-linearly related. In experiments, we evaluate the effectiveness of our method using artificial data generated by non-linear causal structures and real data.} }
Endnote
%0 Conference Paper %T Estimating individual-level optimal causal interventions combining causal models and machine learning models %A Keisuke Kiritoshi %A Tomonori Izumitani %A Kazuki Koyama %A Tomomi Okawachi %A Keisuke Asahara %A Shohei Shimizu %B Proceedings of The KDD'21 Workshop on Causal Discovery %C Proceedings of Machine Learning Research %D 2021 %E Thuc Duy Le %E Jiuyong Li %E Greg Cooper %E Sofia Triantafyllou %E Elias Bareinboim %E Huan Liu %E Negar Kiyavash %F pmlr-v150-kiritoshi21a %I PMLR %P 55--77 %U https://proceedings.mlr.press/v150/kiritoshi21a.html %V 150 %X We introduce a new statistical causal inference method to estimate individual-level optimal causal intervention, that is, to which value we should set the value of a certain variable of an individual to obtain a desired value of another variable. This is defined as an optimization problem to minimize the error between a desired value and the value that would have been attained under the setting for the individual. To solve the optimization problem, we first train a machine learning model to predict the value of an objective variable and then estimate the causal structure of variables. We then combine the machine learning model and causal structure into a single causal model to estimate counterfactual value of the predicted objective variable. This is effective in achieving a more accurate estimation of individual-level optimal causal intervention. We further propose a gradient descent algorithm to compute the optimal causal intervention. Our method is generally applicable to continuous variables that are linearly and non-linearly related. In experiments, we evaluate the effectiveness of our method using artificial data generated by non-linear causal structures and real data.
APA
Kiritoshi, K., Izumitani, T., Koyama, K., Okawachi, T., Asahara, K. & Shimizu, S.. (2021). Estimating individual-level optimal causal interventions combining causal models and machine learning models. Proceedings of The KDD'21 Workshop on Causal Discovery, in Proceedings of Machine Learning Research 150:55-77 Available from https://proceedings.mlr.press/v150/kiritoshi21a.html.

Related Material