More efficient policy learning via optimal retargeting
From MaRDI portal
Abstract: Policy learning can be used to extract individualized treatment regimes from observational data in healthcare, civics, e-commerce, and beyond. One big hurdle to policy learning is a commonplace lack of overlap in the data for different actions, which can lead to unwieldy policy evaluation and poorly performing learned policies. We study a solution to this problem based on retargeting, that is, changing the population on which policies are optimized. We first argue that at the population level, retargeting may induce little to no bias. We then characterize the optimal reference policy and retargeting weights in both binary-action and multi-action settings. We do this in terms of the asymptotic efficient estimation variance of the new learning objective. Extensive empirical results in a simulation study and a case study of personalized job counseling demonstrate that retargeting is a fairly easy way to significantly improve any policy learning procedure applied to observational data.
Recommendations
- Discussion of: ``More efficient policy learning via optimal retargeting and ``Learning optimal distributionally robust individualized treatment rules
- Policy learning with observational data
- Efficient augmentation and relaxation learning for individualized treatment rules using observational data
- Learning when-to-treat policies
- Doubly robust policy evaluation and optimization
Cites work
- Asymptotic Statistics
- Asymptotics for statistical treatment rules
- Balancing covariates via propensity score weighting
- Does matching overcome LaLonde's critique of nonexperimental estimators?
- Double/debiased machine learning for treatment and structural parameters
- Dynamic treatment regimes: technical challenges and applications
- Efficient Estimation of Average Treatment Effects Using the Estimated Propensity Score
- Estimating individualized treatment rules using outcome weighted learning
- Estimation of Regression Coefficients When Some Regressors Are Not Always Observed
- Generalized optimal matching methods for causal inference
- scientific article; zbMATH DE number 51427 (Why is no real title available?)
- scientific article; zbMATH DE number 3456281 (Why is no real title available?)
- scientific article; zbMATH DE number 490141 (Why is no real title available?)
- scientific article; zbMATH DE number 1391397 (Why is no real title available?)
- scientific article; zbMATH DE number 6542809 (Why is no real title available?)
- Matching As An Econometric Evaluation Estimator: Evidence from Evaluating a Job Training Programme
- Minimax regret treatment choice with finite samples
- Multivariate matching methods that are monotonic imbalance bounding
- On the Role of the Propensity Score in Efficient Semiparametric Estimation of Average Treatment Effects
- Optimal probability weights for inference with constrained precision
- Overlap in observational studies with high-dimensional covariates
- Performance guarantees for individualized treatment rules
- Robustifying trial-derived optimal treatment rules for a target population
- Semiparametric theory and missing data.
- Who should be treated? Empirical welfare maximization methods for treatment choice
Cited in
(12)- Constructing effective personalized policies using counterfactual inference from biased data sets with many features
- Discussion of Kallus (2020) and Mo, Qi, and Liu (2020): New Objectives for Policy Learning
- Discussion of: ``More efficient policy learning via optimal retargeting and ``Learning optimal distributionally robust individualized treatment rules
- Rejoinder: New Objectives for Policy Learning
- Policy learning with observational data
- Assumption-Lean Cox Regression
- Offline Multi-Action Policy Learning: Generalization and Optimization
- Estimation of optimal dynamic treatment assignment rules under policy constraints
- Efficient and robust transfer learning of optimal individualized treatment regimes with right-censored survival data
- Parameterizing the effect of a continuous treatment using average derivative effects
- On weighted orthogonal learners for heterogeneous treatment effects
- Binary choice under asymmetric loss in a data-rich environment: theory and an application to algorithmic fairness
This page was built for publication: More efficient policy learning via optimal retargeting
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4999139)
