# Proximal Learning for Individualized Treatment Regimes Under Unmeasured Confounding

@article{Qi2021ProximalLF, title={Proximal Learning for Individualized Treatment Regimes Under Unmeasured Confounding}, author={Zhengling Qi and Rui Miao and Xiaoke Zhang}, journal={ArXiv}, year={2021}, volume={abs/2105.01187} }

Data-driven individualized decision making has recently received increasing research interests. Most existing methods rely on the assumption of no unmeasured confounding, which unfortunately cannot be ensured in practice especially in observational studies. Motivated by the recent proposed proximal causal inference, we develop several proximal learning approaches to estimating optimal individualized treatment regimes (ITRs) in the presence of unmeasured confounding. In particular, we establish…

## 11 Citations

### Proximal Causal Inference for Marginal Counterfactual Survival Curves

- Mathematics
- 2022

Contrasting marginal counterfactual survival curves across treatment arms is an eﬀective and popular approach for inferring the causal eﬀect of an intervention on a right-censored time-to-event…

### A non-parametric Bayesian approach for adjusting partial compliance in sequential decision making

- Mathematics
- 2021

Existing methods in estimating the mean outcome under a given dynamic treatment regime rely on intention-to-treat analyses which estimate the effect of following a certain dynamic treatment regime…

### RISE: Robust Individualized Decision Learning with Sensitive Variables

- Computer Science
- 2022

This paper introduces RISE, a robust individualized decision learning framework with sensitive variables, where sensitive variables are collectible data and important to the intervention decision,…

### Long-term Causal Inference Under Persistent Confounding via Data Combination

- Mathematics, Economics
- 2022

We study the identiﬁcation and estimation of long-term treatment eﬀects when both experimental and observational data are available. Since the long-term outcome is observed only after a long delay,…

### Proximal Identification and Estimation to Handle Dependent Right Censoring for Survival Analysis

- Mathematics
- 2022

Modern epidemiological and clinical studies aim at analyzing a time-to-event endpoint. A common complication is right censoring. In some cases, it arises because subjects are still surviving after…

### Debiased Inference on Identified Linear Functionals of Underidentified Nuisances via Penalized Minimax Estimation

- Mathematics
- 2022

We study generic inference on identiﬁed linear functionals of nonunique nuisances deﬁned as solutions to underidentiﬁed conditional moment restrictions. This problem appears in a variety of…

### Controlling for Unmeasured Confounding in Panel Data Using Minimal Bridge Functions: From Two-Way Fixed Effects to Factor Models

- Mathematics, Economics
- 2021

It is shown that analogous, albeit more complex, transformations exist in the more general linear factor model, providing a new means to identify the effect in that model, and is proved that the resulting average causal effect estimator is root-N consistent and asymptotically normal.

### Double Negative Control Inference in Test-Negative Design Studies of Vaccine Effectiveness

- MedicineArXiv
- 2022

A novel approach is presented to identify and estimate vaccine effectiveness in the target population by carefully leveraging a pair of negative control exposure and outcome variables to account for potential hidden bias in TND studies.

### Optimal regimes for algorithm-assisted human decision-making

- Economics, Computer Science
- 2022

This work introduces optimal regimes for algorithm-assisted human decision-making that enjoy a “superoptimality” property whereby they are guaranteed to outperform conventional optimal regimes currently considered in the literature.

### Rate-Optimal Contextual Online Matching Bandit

- Computer ScienceArXiv
- 2022

A Centralized Contextual Explore Then Commit (CC-ETC) algorithm to adapt to the COMBO, which solves online matching with dynamic preference and achieves a sublinear regret upper bound O(log(T )) and is a rate-optimal algorithm by proving a matching lower bound.

## References

SHOWING 1-10 OF 61 REFERENCES

### A Semiparametric Instrumental Variable Approach to Optimal Treatment Regimes Under Endogeneity

- EconomicsJournal of the American Statistical Association
- 2021

A general instrumental variable approach to learning optimal treatment regimes under endogeneity is proposed and establishes identification of both value function for a given regime and optimal regimes with the aid of a binary IV, when no unmeasured confounding fails to hold.

### Estimating Optimal Treatment Rules with an Instrumental Variable: A Semi-Supervised Learning Approach

- Mathematics, Computer Science
- 2020

A statistical learning method is proposed that estimates such an IV-optimal rule, design computationally-efficient algorithms, and prove theoretical guarantees and applies the method to studying which moms would better deliver their premature babies at hospitals with high-level neonatal intensive care units (NICUs).

### Semiparametric proximal causal inference

- Mathematics
- 2020

This paper considers the framework of proximal causal inference introduced by Tchetgen Tchet Gen et al. (2020), which while explicitly acknowledging covariate measurements as imperfect proxies of confounding mechanisms, offers an opportunity to learn about causal effects in settings where exchangeability on the basis of measured covariates fails.

### Optimal Individualized Decision Rules Using Instrumental Variable Methods

- EconomicsJournal of the American Statistical Association
- 2021

Methods to estimate optimal individualized rules and construct asymptotically efficient plug-in estimators of the corresponding average causal effect relative to a prespecified reference rule are developed.

### Efficient augmentation and relaxation learning for individualized treatment rules using observational data

- EconomicsJ. Mach. Learn. Res.
- 2019

This work considers a class of estimators for optimal treatment rules that are analogous to convex large-margin classifiers and derives rates of convergence for the proposed estimators and uses these rates to characterize the bias-variance trade-off for estimating individualized treatment rules with classification-based methods.

### Policy Learning With Observational Data

- Economics, Mathematics
- 2017

Given a doubly robust estimator of the causal effect of assigning everyone to treatment, an algorithm for choosing whom to treat is developed, and strong guarantees for the asymptotic utilitarian regret of the resulting policy are established.

### An Introduction to Proximal Causal Learning

- MathematicsmedRxiv
- 2020

A formal potential outcome framework for proximal causal learning is introduced, which while explicitly acknowledging covariate measurements as imperfect proxies of confounding mechanisms, offers an opportunity to learn about causal effects in settings where exchangeability on the basis of measured covariates fails.

### Causal Inference Under Unmeasured Confounding With Negative Controls: A Minimax Learning Approach

- MathematicsArXiv
- 2021

This paper tackles the primary challenge to causal inference using negative controls: the identification and estimation of these bridge functions, and provides a new identification strategy that avoids both uniqueness and completeness.

### Multiply robust causal inference with double‐negative control adjustment for categorical unmeasured confounding

- MathematicsJournal of the Royal Statistical Society. Series B, Statistical methodology
- 2020

This work establishes non-parametric identification of the ATE under weaker conditions in the case of categorical unmeasured confounding and negative control variables, and provides a general semiparametric framework for obtaining inferences about the AtE while leveraging information about a possibly large number of measured covariates.

### Quantile-Optimal Treatment Regimes

- Mathematics, EconomicsJournal of the American Statistical Association
- 2018

This article proposes an alternative formulation of the estimator as a solution of an optimization problem with an estimated nuisance parameter, and derives theory involving a nonstandard convergence rate and a nonnormal limiting distribution of the quantile-optimal treatment regime.