mlsanityData to 14 Oct 2026

Learning Deployable Causal Action Geometry under Temporal Non-Stationarity

arXiv 2610.16644 ↗First posted 12 Oct 2026cs.LGcited by 0

Not on the map. Why?

Abstract (arXiv)

Learning a response law for continuous decisions from observational time series is central to many real-world decision systems, yet temporal non-stationarity poses a fundamental challenge: as outcomes drift over time, naively pooled models can conflate action effects with shifting baselines and even reverse action rankings. Even when the one-step response is identifiable, unrestricted cross-time generalization remains fundamentally ill-posed. We therefore seek the part of the response that remains stable for decisions even when outcome levels do not. We propose deployable causal action geometry: the action-response shape that remains after anchoring away action-independent level shifts on a selected link scale and profiling out positive amplitude variation. Building on this target, we develop a two-stage orthogonal estimator that first obtains cross-fitted pilot estimates of anchored responses and then performs honest profiled selection over link, representation, and shape classes. The resulting estimator admits oracle guarantees for the restricted target and, under approximate persistence, produces decision rules that remain near optimal in future periods. Experiments on synthetic data and Dominick's support the proposed target and its transfer behavior. In a randomized Taobao A/B test, the deployed system improves GMV and purchase counts by (1.16%) and (1.70%).

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 22 of its 31 references that are in the map. First posted on arXiv 12 Oct 2026.

Builds on

22 of 31 references.

References · 22

  1. 2015Domain-Adversarial Training of Neural Networkscited by 3,044
  2. 2019Invariant Risk Minimizationcited by 1,854
  3. 2020Out-of-Distribution Generalization via Risk Extrapolation (REx)cited by 742
  4. 2015Causal inference using invariant prediction: identification and confidence intervalscited by 536
  5. 2020Estimating Counterfactual Treatment Outcomes over Time Through Adversarially Balanced Representationscited by 155
  6. 2019Orthogonal Statistical Learningcited by 151
  7. 2017Domain Adaptation by Using Causal Inference to Predict Invariant Conditional Distributionscited by 126
  8. 2022Causal Transformer for Estimating Counterfactual Outcomescited by 123
  9. 2018Anchor regression: heterogeneous data meets causalitycited by 118
  10. 2019Time Series Deconfounder: Estimating Treatment Effects over Time in the Presence of Hidden Confounderscited by 75
  11. 2015Nonparametric methods for doubly robust estimation of continuous treatment effectscited by 64
  12. 2021VCNet and Functional Targeted Regularization For Learning Causal Effects of Continuous Treatmentscited by 61
10 more
  1. 2018Policy Evaluation and Optimization with Continuous Treatmentscited by 58
  2. 2024Model-agnostic meta-learners for estimating heterogeneous treatment effects over timecited by 19
  3. 2020Doubly Robust Off-Policy Value and Gradient Estimation for Deterministic Policiescited by 10
  4. 2021Markdowns in E-Commerce Fresh Retail: A Counterfactual Prediction and Multi-Period Optimization Approachcited by 9
  5. 2025Treatment Effect Estimation for Optimal Decision-Makingcited by 8
  6. 2020Causal Transfer Random Forest: Combining Logged Data and Randomized Experiments for Robust Predictioncited by 5
  7. 2023Causal Forecasting for Pricingcited by 2
  8. 2024Training and Evaluating Causal Forecasting Models for Time-Seriescited by 2
  9. 2026OpFlow: Learning Opportunity-Conditioned Choice Potentials for Robust OD Flow Predictioncited by 1
  10. 2026Multi-channel Uplift Policy Learningcited by 1

Built on by

Not cited yet.