Causal Isotonic Calibration for Heterogeneous Treatment Effects

Lars Van Der Laan, Ernesto Ulloa-Perez, Marco Carone, Alex Luedtke
Proceedings of the 40th International Conference on Machine Learning, PMLR 202:34831-34854, 2023.

Abstract

We propose causal isotonic calibration, a novel nonparametric method for calibrating predictors of heterogeneous treatment effects. Furthermore, we introduce cross-calibration, a data-efficient variant of calibration that eliminates the need for hold-out calibration sets. Cross-calibration leverages cross-fitted predictors and generates a single calibrated predictor using all available data. Under weak conditions that do not assume monotonicity, we establish that both causal isotonic calibration and cross-calibration achieve fast doubly-robust calibration rates, as long as either the propensity score or outcome regression is estimated accurately in a suitable sense. The proposed causal isotonic calibrator can be wrapped around any black-box learning algorithm, providing robust and distribution-free calibration guarantees while preserving predictive performance.

Cite this Paper


BibTeX
@InProceedings{pmlr-v202-van-der-laan23a, title = {Causal Isotonic Calibration for Heterogeneous Treatment Effects}, author = {Van Der Laan, Lars and Ulloa-Perez, Ernesto and Carone, Marco and Luedtke, Alex}, booktitle = {Proceedings of the 40th International Conference on Machine Learning}, pages = {34831--34854}, year = {2023}, editor = {Krause, Andreas and Brunskill, Emma and Cho, Kyunghyun and Engelhardt, Barbara and Sabato, Sivan and Scarlett, Jonathan}, volume = {202}, series = {Proceedings of Machine Learning Research}, month = {23--29 Jul}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v202/van-der-laan23a/van-der-laan23a.pdf}, url = {https://proceedings.mlr.press/v202/van-der-laan23a.html}, abstract = {We propose causal isotonic calibration, a novel nonparametric method for calibrating predictors of heterogeneous treatment effects. Furthermore, we introduce cross-calibration, a data-efficient variant of calibration that eliminates the need for hold-out calibration sets. Cross-calibration leverages cross-fitted predictors and generates a single calibrated predictor using all available data. Under weak conditions that do not assume monotonicity, we establish that both causal isotonic calibration and cross-calibration achieve fast doubly-robust calibration rates, as long as either the propensity score or outcome regression is estimated accurately in a suitable sense. The proposed causal isotonic calibrator can be wrapped around any black-box learning algorithm, providing robust and distribution-free calibration guarantees while preserving predictive performance.} }
Endnote
%0 Conference Paper %T Causal Isotonic Calibration for Heterogeneous Treatment Effects %A Lars Van Der Laan %A Ernesto Ulloa-Perez %A Marco Carone %A Alex Luedtke %B Proceedings of the 40th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2023 %E Andreas Krause %E Emma Brunskill %E Kyunghyun Cho %E Barbara Engelhardt %E Sivan Sabato %E Jonathan Scarlett %F pmlr-v202-van-der-laan23a %I PMLR %P 34831--34854 %U https://proceedings.mlr.press/v202/van-der-laan23a.html %V 202 %X We propose causal isotonic calibration, a novel nonparametric method for calibrating predictors of heterogeneous treatment effects. Furthermore, we introduce cross-calibration, a data-efficient variant of calibration that eliminates the need for hold-out calibration sets. Cross-calibration leverages cross-fitted predictors and generates a single calibrated predictor using all available data. Under weak conditions that do not assume monotonicity, we establish that both causal isotonic calibration and cross-calibration achieve fast doubly-robust calibration rates, as long as either the propensity score or outcome regression is estimated accurately in a suitable sense. The proposed causal isotonic calibrator can be wrapped around any black-box learning algorithm, providing robust and distribution-free calibration guarantees while preserving predictive performance.
APA
Van Der Laan, L., Ulloa-Perez, E., Carone, M. & Luedtke, A.. (2023). Causal Isotonic Calibration for Heterogeneous Treatment Effects. Proceedings of the 40th International Conference on Machine Learning, in Proceedings of Machine Learning Research 202:34831-34854 Available from https://proceedings.mlr.press/v202/van-der-laan23a.html.

Related Material