Stabilized Neural Prediction of Potential Outcomes in Continuous Time
Konstantin Hess, Stefan Feuerriegel
Abstract
Patient trajectories from electronic health records are widely used to estimate conditional average potential outcomes (CAPOs) of treatments over time, which then allows to personalize care. Yet, existing neural methods for this purpose have a key limitation: while some adjust for time-varying confounding, these methods assume that the time series are recorded in discrete time. In other words, they are constrained to settings where measurements and treatments are conducted at fixed time steps, even though this is unrealistic in medical practice. In this work, we aim to estimate CAPOs in continuous time. The latter is of direct practical relevance because it allows for modeling patient trajectories where measurements and treatments take place at arbitrary, irregular timestamps. We thus propose a new method called stabilized continuous time inverse propensity network (SCIP-Net). For this, we further derive stabilized inverse propensity weights for robust estimation of the CAPOs. To the best of our knowledge, our SCIP-Net is the first neural method that performs proper adjustments for time-varying confounding in continuous time.
Ask Lune
Know this paper in minutes, not an afternoon.
Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.
4 top-tier papers have cited it since. Know what they found before your reviewers do.
Lune can draw on
the full paper · the top-tier papers that cite it · related work
Free to start. No credit card required.
Lineage
- Neural SDEs as Infinite-Dimensional GANs cites Neural Controlled Differential Equations for Irregular Time Series
- ContiFormer: Continuous-Time Transformer for Irregular Time Series Modeling cites Neural Controlled Differential Equations for Irregular Time Series
- Causal Transformer for Estimating Counterfactual Outcomes cites Estimating counterfactual treatment outcomes over time through adversarially balanced representations
- Model-agnostic meta-learners for estimating heterogeneous treatment effects over time cites Causal Transformer for Estimating Counterfactual Outcomes
- Model-agnostic meta-learners for estimating heterogeneous treatment effects over time cites Estimating counterfactual treatment outcomes over time through adversarially balanced representations
- IGC-Net for conditional average potential outcome estimation over time cites Causal Transformer for Estimating Counterfactual Outcomes
- IGC-Net for conditional average potential outcome estimation over time cites Estimating counterfactual treatment outcomes over time through adversarially balanced representations
- IGC-Net for conditional average potential outcome estimation over time cites Model-agnostic meta-learners for estimating heterogeneous treatment effects over time
- IGC-Net for conditional average potential outcome estimation over time cites Overlap-weighted orthogonal meta-learner for treatment effect estimation over time
- Overlap-weighted orthogonal meta-learner for treatment effect estimation over time cites IGC-Net for conditional average potential outcome estimation over time
- IGC-Net for conditional average potential outcome estimation over time cites Neural Controlled Differential Equations for Irregular Time Series
- Efficient and Sharp Off-Policy Learning under Unobserved Confounding cites Overlap-weighted orthogonal meta-learner for treatment effect estimation over time
- Overlap-weighted orthogonal meta-learner for treatment effect estimation over time cites Causal Transformer for Estimating Counterfactual Outcomes
- Overlap-weighted orthogonal meta-learner for treatment effect estimation over time cites Estimating counterfactual treatment outcomes over time through adversarially balanced representations
- Overlap-weighted orthogonal meta-learner for treatment effect estimation over time cites Model-agnostic meta-learners for estimating heterogeneous treatment effects over time
Cited by top-tier papers4
- IGC-Net for conditional average potential outcome estimation over timeKonstantin Hess, Dennis Frauen, Valentyn Melnychuk, Stefan FeuerriegelICLR 2026 · 8 citations
- Overlap-weighted orthogonal meta-learner for treatment effect estimation over timeKonstantin Hess, Dennis Frauen, Mihaela van der Schaar, Stefan FeuerriegelICLR 2026 · 7 citations
- Efficient and Sharp Off-Policy Learning under Unobserved ConfoundingKonstantin Hess, Dennis Frauen, Valentyn Melnychuk, Stefan FeuerriegelICLR 2026 · 5 citations
- Model-agnostic meta-learners for estimating heterogeneous treatment effects over timeDennis Frauen, Konstantin Hess, Stefan FeuerriegelICLR 2025
Builds on14
- Neural Controlled Differential Equations for Irregular Time SeriesPatrick Kidger, James Morrill, James Foster, Terry J. LyonsNeurIPS 2020 · 850 citations
- Estimating counterfactual treatment outcomes over time through adversarially balanced representationsIoana Bica, Ahmed M. Alaa, James Jordon, Mihaela van der SchaarICLR 2020 · 224 citations
- Neural SDEs as Infinite-Dimensional GANsPatrick Kidger, James Foster, Xuechen Li, Terry J. LyonsICML 2021 · 214 citations
- Causal Transformer for Estimating Counterfactual OutcomesValentyn Melnychuk, Dennis Frauen, Stefan FeuerriegelICML 2022 · 146 citations
- ContiFormer: Continuous-Time Transformer for Irregular Time Series ModelingYuqi Chen, Kan Ren, Yansen Wang, Yuchen Fang et al.NeurIPS 2023 · 131 citations
Related papers
- Continuous-Time Modeling of Counterfactual Outcomes Using Neural Controlled Differential EquationsNabeel Seedat, Fergus Imrie, Alexis Bellot, Zhaozhi Qian et al.ICML 2022 · 68 citations
- Bayesian Neural Controlled Differential Equations for Treatment Effect EstimationKonstantin Hess, Valentyn Melnychuk, Dennis Frauen, Stefan FeuerriegelICLR 2024 · 27 citations
- VCNet and Functional Targeted Regularization For Learning Causal Effects of Continuous TreatmentsLizhen Nie, Mao Ye, Qiang Liu, Dan NicolaeICLR 2021 · 81 citations
- Conformal Prediction for Causal Effects of Continuous TreatmentsMaresa Schröder, Dennis Frauen, Jonas Schweisthal, Konstantin Hess et al.NeurIPS 2025 · 21 citations
- Estimating Treatment Effects from Irregular Time Series Observations with Hidden ConfoundersDefu Cao, James Enouen, Yujing Wang, Xiangchen Song et al.AAAI 2023 · 28 citations
