EconBase
← All papers

Semi-Parametric Efficient Policy Learning with Continuous Actions

Mert Demirer, Vasilis Syrgkanis, Greg Lewis, Victor Chernozhukov

arXiv 24 May 2019 · Econometrics · 16 citations (OpenAlex)

arXiv:1905.10116 · PDF · DOI · OpenAlex · Extracted main text

Abstract

We consider off-policy evaluation and optimization with continuous action spaces. We focus on observational data where the data collection policy is unknown and needs to be estimated. We take a semi-parametric approach where the value function takes a known parametric form in the treatment, but we are agnostic on how it depends on the observed contexts. We propose a doubly robust off-policy estimate for this setting and show that off-policy optimization based on this estimate is robust to estimation errors of the policy function or the regression model. Our results also apply if the model does not satisfy our semi-parametric form, but rather we measure regret in terms of the best projection of the true value function to this functional space. Our work extends prior approaches of policy optimization from observational data that only considered discrete actions. We provide an experimental evaluation of our method in a synthetic data example motivated by optimal personalized pricing and costly resource allocation.

Citation extraction

28
references
59
in-text mentions
28
distinct cited
3
self-citations
6,920
main-text words

appendix boundary found by appendix_command · 44% of the source is main text. Read the extracted text to check this.

Most heavily cited references

The works this paper leans on most, across its whole bibliography — not restricted to papers in our corpus. Ranked by composite intensity, which combines how often a work is mentioned, how many sections mention it, and how much of that falls in the main text rather than the appendix.

ReferenceIntensityMentionsSectionsMain text
1Susan Athey and Stefan Wager (2018) Efficient policy learning0.9507386%
2Dylan J Foster and Vasilis Syrgkanis (2019) Orthogonal statistical learning self0.71411336%
3Adith Swaminathan and Thorsten Joachims (2015) Counterfactual risk minimization: Learning from logged bandit feedback0.69351100%
4Nathan Kallus and Angela Zhou (2018) Policy evaluation and optimization with continuous treatments0.64441100%
5Victor Chernozhukov, Denis Chetverikov, Mert Demirer, Esther Duflo,… (2018) Double/debiased machine learning for treatment and structural parameters self0.64422100%
6Zhengyuan Zhou, Susan Athey, and Stefan Wager (2018) Offline multi-action policy learning: Generalization and optimization0.58531100%
7A. W. Van Der Vaart and J. A. Wellner (1996) Weak Convergence and Empirical Processes: With Applications to Statistics0.5112250%
8Bryan S Graham and Cristine Campos de Xavier Pinto (2018) Semiparametrically efficient estimation of the average linear regression function0.51121100%
9Akshay Krishnamurthy, John Langford, Aleksandrs Slivkins, and Chiche… (2019) Contextual bandits with continuous actions: Smoothing, zooming, and adapting0.51121100%
10Martin J. Wainwright (2019) High-Dimensional Statistics: A Non-Asymptotic Viewpoint0.51121100%

Showing the top 10 of 28 scored citations.

Cited by, within the corpus

arXiv econ.EM papers that cite this one, ranked by how heavily they lean on it.

Citing paperIntensityMentionsSections
1Double Debiased Machine Learning Nonparametric Inference with Continuous Treatments0.73733
2Individualized Policy Evaluation and Learning under Clustered Network Interference0.73732
3Personalized Policy Learning through Discrete Experimentation: Theory and Empirical Evidence0.73732
4Efficient Policy Learning from Surrogate-Loss Classification Reductions0.51121
5Policy Learning with Observational Data : The Case of Hepatitis C Treatment for HIV/HCV Co-Infected Patients0.51121
6Minimax Kernel Machine Learning for a Class of Doubly Robust Functionals with Application to Proximal Causal Inference0.40521
7Off-Policy Evaluation and Learning for External Validity under a Covariate Shift0.40511
8Confidence Interval for Off-Policy Evaluation from Dependent Samples via Bandit Algorithm: Approach from Standardized Martingales0.40511
9Empirical Welfare Maximization with Constraints0.40511
10Policy Learning with Adaptively Collected Data0.40511