mlsanityData to 7 Oct 2026

Stabilized Inverse Probability Weighting via Isotonic Calibration

arXiv 2411.06342 ↗First posted 10 Nov 2024stat.ME stat.MLcited by 8

Not on the map. Why?

Abstract (arXiv)

Inverse weighting with an estimated propensity score is widely used by estimation methods in causal inference to adjust for confounding bias. However, directly inverting propensity score estimates can lead to instability, bias, and excessive variability due to large inverse weights, especially when treatment overlap is limited. In this work, we propose a post-hoc calibration algorithm for inverse propensity weights that generates well-calibrated, stabilized weights from user-supplied, cross-fitted propensity score estimates. Our approach employs a variant of isotonic regression with a loss function specifically tailored to the inverse propensity weights. Through theoretical analysis and empirical studies, we demonstrate that isotonic calibration improves the performance of doubly robust estimators of the average treatment effect.

Key facts. Cited by 8 papers in the map (data through 7 Oct 2026). Builds on 18 of its 53 references that are in the map. First posted on arXiv 10 Nov 2024.

Builds on

18 of 53 references.

References · 18

  1. 2016XGBoost: A Scalable Tree Boosting Systemcited by 5,120
  2. 2017On Calibration of Modern Neural Networkscited by 4,860
  3. 2021Distribution-free calibration guarantees for histogram binning without sample splittingcited by 44
  4. 2021RieszNet and ForestRiesz: Automatic Debiased Machine Learning with Neural Nets and Random Forestscited by 42
  5. 2016Covariate Balancing Propensity Score by Tailored Loss Functionscited by 34
  6. 2017Augmented Minimax Linear Estimationcited by 33
  7. 2019Atlantic Causal Inference Conference (ACIC) Data Analysis Challenge 2017cited by 23
  8. 2023Augmented balancing weights as linear regressioncited by 21
  9. 2021The Balancing Act in Causal Inferencecited by 20
  10. 2017Minimal Dispersion Approximately Balancing Weights: Asymptotic Properties and Practical Considerationscited by 18
  11. 2023Causal isotonic calibration for heterogeneous treatment effectscited by 16
  12. 2024Self-Calibrating Conformal Predictioncited by 16
6 more
  1. 2024Combining T-learning and DR-learning: a framework for oracle-efficient estimation of causal contrastscited by 15
  2. 2019Minimax Linear Estimation of the Retargeted Meancited by 11
  3. 2020Propensity score weighting under limited overlap and model misspecificationcited by 8
  4. 2024Doubly robust inference via calibrationcited by 8
  5. 2017On Adaptive Propensity Score Truncation in Causal Inferencecited by 6
  6. 2022Isotonic propensity score matchingcited by 1

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 5; 2026: 3.

8 papers.

  1. 2025Generalized Venn and Venn-Abers Calibration with Applications in Conformal Predictioncited by 6
  2. 2025Automatic Debiased Machine Learning for Smooth Functionals of Nonparametric M-Estimandscited by 5
  3. 2025Causal Judge Evaluation: Calibrated Surrogate Metrics for LLM Systemscited by 2
  4. 2025Bellman Calibration for V-Learning in Offline Reinforcement Learningcited by 2
  5. 2025Text Rationalization for Robust Causal Effect Estimationcited by 1
  6. 2026Tailoring Strictly Proper Scoring Rules for Downstream Tasks: An Application to Causal Inferencecited by 1
  7. 2026Data-Poisoning Audits for Causal Effect Estimationcited by 0
  8. 2026Bellman Calibration for Marginalized Importance Weighting in Offline Reinforcement Learningcited by 0