Estimating Interventional Outcomes over Time with Causal Normalizing Flow

Yoonseok Yeom, Jonghwan Kim, Taehui Yun, Juhyun Lyu, Jung-Hee Kim, Sangmin Lee, Jinseok Yang, Hyemin Jung, Woohyung Lim, Sanghack Lee
Proceedings of the 42nd Conference on Uncertainty in Artificial Intelligence, PMLR 337:7856-7895, 2026.

Abstract

Estimating outcome distributions under time-varying treatments is an essential task for personalized decision-making, particularly in domains such as healthcare. Most prior work in this area focuses on point predictions, which fail to capture the inherent variability in outcomes. Recent efforts in causal inference have begun integrating generative models to address this limitation by estimating interventional distributions. However, existing approaches—including causal normalizing flows—are generally restricted to static settings and are not well suited to sequential, time-dependent data. In this work, we propose a novel framework that extends causal normalizing flows to time-series, enabling simulation-based interventional density estimation over time. Our method learns representations of treatment and covariate history that capture temporal dependencies. Conditioned on these representations and guided by a causal graph, our flow-based model generates interventional samples, allowing for the simulation of outcome trajectories under alternative treatment strategies. We evaluate our approach on both linear and non-linear synthetic time-series as well as on a simulated tumor growth dataset, demonstrating that it achieves performance competitive with state-of-the-art baselines, while accommodating a broader spectrum of causal queries.

Cite this Paper

BibTeX
@InProceedings{pmlr-v337-yeom26a, title = {Estimating Interventional Outcomes over Time with Causal Normalizing Flow}, author = {Yeom, Yoonseok and Kim, Jonghwan and Yun, Taehui and Lyu, Juhyun and Kim, Jung-Hee and Lee, Sangmin and Yang, Jinseok and Jung, Hyemin and Lim, Woohyung and Lee, Sanghack}, booktitle = {Proceedings of the 42nd Conference on Uncertainty in Artificial Intelligence}, pages = {7856--7895}, year = {2026}, editor = {Perković, Emilija and Malinsky, Daniel}, volume = {337}, series = {Proceedings of Machine Learning Research}, month = {17--21 Aug}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v337/main/assets/yeom26a/yeom26a.pdf}, url = {https://proceedings.mlr.press/v337/yeom26a.html}, abstract = {Estimating outcome distributions under time-varying treatments is an essential task for personalized decision-making, particularly in domains such as healthcare. Most prior work in this area focuses on point predictions, which fail to capture the inherent variability in outcomes. Recent efforts in causal inference have begun integrating generative models to address this limitation by estimating interventional distributions. However, existing approaches—including causal normalizing flows—are generally restricted to static settings and are not well suited to sequential, time-dependent data. In this work, we propose a novel framework that extends causal normalizing flows to time-series, enabling simulation-based interventional density estimation over time. Our method learns representations of treatment and covariate history that capture temporal dependencies. Conditioned on these representations and guided by a causal graph, our flow-based model generates interventional samples, allowing for the simulation of outcome trajectories under alternative treatment strategies. We evaluate our approach on both linear and non-linear synthetic time-series as well as on a simulated tumor growth dataset, demonstrating that it achieves performance competitive with state-of-the-art baselines, while accommodating a broader spectrum of causal queries.} }
Endnote
%0 Conference Paper %T Estimating Interventional Outcomes over Time with Causal Normalizing Flow %A Yoonseok Yeom %A Jonghwan Kim %A Taehui Yun %A Juhyun Lyu %A Jung-Hee Kim %A Sangmin Lee %A Jinseok Yang %A Hyemin Jung %A Woohyung Lim %A Sanghack Lee %B Proceedings of the 42nd Conference on Uncertainty in Artificial Intelligence %C Proceedings of Machine Learning Research %D 2026 %E Emilija Perković %E Daniel Malinsky %F pmlr-v337-yeom26a %I PMLR %P 7856--7895 %U https://proceedings.mlr.press/v337/yeom26a.html %V 337 %X Estimating outcome distributions under time-varying treatments is an essential task for personalized decision-making, particularly in domains such as healthcare. Most prior work in this area focuses on point predictions, which fail to capture the inherent variability in outcomes. Recent efforts in causal inference have begun integrating generative models to address this limitation by estimating interventional distributions. However, existing approaches—including causal normalizing flows—are generally restricted to static settings and are not well suited to sequential, time-dependent data. In this work, we propose a novel framework that extends causal normalizing flows to time-series, enabling simulation-based interventional density estimation over time. Our method learns representations of treatment and covariate history that capture temporal dependencies. Conditioned on these representations and guided by a causal graph, our flow-based model generates interventional samples, allowing for the simulation of outcome trajectories under alternative treatment strategies. We evaluate our approach on both linear and non-linear synthetic time-series as well as on a simulated tumor growth dataset, demonstrating that it achieves performance competitive with state-of-the-art baselines, while accommodating a broader spectrum of causal queries.
APA
Yeom, Y., Kim, J., Yun, T., Lyu, J., Kim, J., Lee, S., Yang, J., Jung, H., Lim, W. & Lee, S.. (2026). Estimating Interventional Outcomes over Time with Causal Normalizing Flow. Proceedings of the 42nd Conference on Uncertainty in Artificial Intelligence, in Proceedings of Machine Learning Research 337:7856-7895 Available from https://proceedings.mlr.press/v337/yeom26a.html.

Related Material