IDEAS home Printed from https://ideas.repec.org/p/arx/papers/2409.00379.html
   My bibliography  Save this paper

Bandit Algorithms for Policy Learning: Methods, Implementation, and Welfare-performance

Author

Listed:
  • Toru Kitagawa
  • Jeff Rowley

Abstract

Static supervised learning-in which experimental data serves as a training sample for the estimation of an optimal treatment assignment policy-is a commonly assumed framework of policy learning. An arguably more realistic but challenging scenario is a dynamic setting in which the planner performs experimentation and exploitation simultaneously with subjects that arrive sequentially. This paper studies bandit algorithms for learning an optimal individualised treatment assignment policy. Specifically, we study applicability of the EXP4.P (Exponential weighting for Exploration and Exploitation with Experts) algorithm developed by Beygelzimer et al. (2011) to policy learning. Assuming that the class of policies has a finite Vapnik-Chervonenkis dimension and that the number of subjects to be allocated is known, we present a high probability welfare-regret bound of the algorithm. To implement the algorithm, we use an incremental enumeration algorithm for hyperplane arrangements. We perform extensive numerical analysis to assess the algorithm's sensitivity to its tuning parameters and its welfare-regret performance. Further simulation exercises are calibrated to the National Job Training Partnership Act (JTPA) Study sample to determine how the algorithm performs when applied to economic data. Our findings highlight various computational challenges and suggest that the limited welfare gain from the algorithm is due to substantial heterogeneity in causal effects in the JTPA data.

Suggested Citation

  • Toru Kitagawa & Jeff Rowley, 2024. "Bandit Algorithms for Policy Learning: Methods, Implementation, and Welfare-performance," Papers 2409.00379, arXiv.org.
  • Handle: RePEc:arx:papers:2409.00379
    as

    Download full text from publisher

    File URL: http://arxiv.org/pdf/2409.00379
    File Function: Latest version
    Download Restriction: no
    ---><---

    References listed on IDEAS

    as
    1. Alberto Abadie & Joshua Angrist & Guido Imbens, 2002. "Instrumental Variables Estimates of the Effect of Subsidized Training on the Quantiles of Trainee Earnings," Econometrica, Econometric Society, vol. 70(1), pages 91-117, January.
    2. Maximilian Kasy & Anja Sautmann, 2021. "Adaptive Treatment Assignment in Experiments for Policy Choice," Econometrica, Econometric Society, vol. 89(1), pages 113-132, January.
    3. Susan Athey & Raj Chetty & Guido W. Imbens & Hyunseung Kang, 2019. "The Surrogate Index: Combining Short-Term Proxies to Estimate Long-Term Treatment Effects More Rapidly and Precisely," NBER Working Papers 26463, National Bureau of Economic Research, Inc.
    4. Howard S. Bloom & Larry L. Orr & Stephen H. Bell & George Cave & Fred Doolittle & Winston Lin & Johannes M. Bos, 1997. "The Benefits and Costs of JTPA Title II-A Programs: Key Findings from the National Job Training Partnership Act Study," Journal of Human Resources, University of Wisconsin Press, vol. 32(3), pages 549-576.
    5. Susan Athey & Undral Byambadalai & Vitor Hadad & Sanath Kumar Krishnamurthy & Weiwen Leung & Joseph Jay Williams, 2022. "Contextual Bandits in a Survey Experiment on Charitable Giving: Within-Experiment Outcomes versus Policy Learning," Papers 2211.12004, arXiv.org.
    Full references (including those not matched with items on IDEAS)

    Citations

    Citations are extracted by the CitEc Project, subscribe to its RSS feed for this item.
    as


    Cited by:

    1. Timothy Armstrong & Martin Weidner & Andrei Zeleneev, 2024. "Robust estimation and inference in panels with interactive fixed effects," IFS Working Papers WCWP28/24, Institute for Fiscal Studies.
    2. Toru Kitagawa & Weining Wang & Mengshan Xu, 2024. "Policy choice in time series by empirical welfare maximization," CeMMAP working papers 27/24, Institute for Fiscal Studies.

    Most related items

    These are the items that most often cite the same works as this one and are cited by the same works as this one.
    1. Toru Kitagawa & Jeff Rowley, 2024. "Bandit algorithms for policy learning: methods, implementation, and welfare-performance," The Japanese Economic Review, Springer, vol. 75(3), pages 407-447, July.
    2. Alan B. Krueger, 2002. "Inequality, Too Much of a Good Thing," Working Papers 845, Princeton University, Department of Economics, Industrial Relations Section..
    3. Undral Byambadalai, 2022. "Identification and Inference for Welfare Gains without Unconfoundedness," Papers 2207.04314, arXiv.org.
    4. Markus Frölich & Blaise Melly, 2013. "Identification of Treatment Effects on the Treated with One-Sided Non-Compliance," Econometric Reviews, Taylor & Francis Journals, vol. 32(3), pages 384-414, November.
    5. A Stefano Caria & Grant Gordon & Maximilian Kasy & Simon Quinn & Soha Osman Shami & Alexander Teytelboym, 2024. "An Adaptive Targeted Field Experiment: Job Search Assistance for Refugees in Jordan," Journal of the European Economic Association, European Economic Association, vol. 22(2), pages 781-836.
    6. repec:pri:cepsud:87krueger is not listed on IDEAS
    7. Toru Kitagawa & Hugo Lopez & Jeff Rowley, 2022. "Stochastic Treatment Choice with Empirical Welfare Updating," Papers 2211.01537, arXiv.org, revised Feb 2023.
    8. Agata Maida & Daniela Sonedda, 2019. "Getting out of the starting gate on the right foot: employment effects of investment in human capital," LABORatorio R. Revelli Working Papers Series 164, LABORatorio R. Revelli, Centre for Employment Studies.
    9. Battistin, Erich & De Nadai, Michele & Krishnan, Nandini, 2023. "The insights and illusions of consumption measurements," Journal of Development Economics, Elsevier, vol. 161(C).
    10. Alan B. Krueger, 2002. "Inequality, Too Much of a Good Thing," Working Papers 115, Princeton University, Department of Economics, Center for Economic Policy Studies..
    11. Toru Kitagawa & Aleksey Tetenov, 2018. "Who Should Be Treated? Empirical Welfare Maximization Methods for Treatment Choice," Econometrica, Econometric Society, vol. 86(2), pages 591-616, March.
    12. Toru Kitagawa & Aleksey Tetenov, 2017. "Equality-minded treatment choice," CeMMAP working papers 10/17, Institute for Fiscal Studies.
    13. Kaspar Wüthrich, 2020. "A Comparison of Two Quantile Models With Endogeneity," Journal of Business & Economic Statistics, Taylor & Francis Journals, vol. 38(2), pages 443-456, April.
    14. Toru Kitagawa & Aleksey Tetenov, 2021. "Equality-Minded Treatment Choice," Journal of Business & Economic Statistics, Taylor & Francis Journals, vol. 39(2), pages 561-574, March.
    15. Gilles Crommen & Jad Beyhum & Ingrid Van Keilegom, 2024. "An instrumental variable approach under dependent censoring," TEST: An Official Journal of the Spanish Society of Statistics and Operations Research, Springer;Sociedad de Estadística e Investigación Operativa, vol. 33(2), pages 473-495, June.
    16. Hsu, Yu-Chin & Huang, Ta-Cheng & Xu, Haiqing, 2023. "Testing For Unobserved Heterogeneous Treatment Effects With Observational Data," Econometric Theory, Cambridge University Press, vol. 39(3), pages 582-622, June.
    17. Toru Kitagawa & Aleksey Tetenov, 2015. "Who should be treated? Empirical welfare maximization methods for treatment choice," CeMMAP working papers 10/15, Institute for Fiscal Studies.
    18. Alan Krueger, 2002. "Inequality, Too Much of a Good Thing," Working Papers 845, Princeton University, Department of Economics, Industrial Relations Section..
    19. Kirill Ponomarev & Vira Semenova, 2024. "On the Lower Confidence Band for the Optimal Welfare," Papers 2410.07443, arXiv.org, revised Oct 2024.
    20. Christophe Bruneel-Zupanc & Jad Beyhum, 2024. "Identification with possibly invalid IVs," Papers 2401.03990, arXiv.org, revised Oct 2024.
    21. Bera Anil K. & Galvao Antonio F. & Montes-Rojas Gabriel V. & Park Sung Y., 2016. "Asymmetric Laplace Regression: Maximum Likelihood, Maximum Entropy and Quantile Regression," Journal of Econometric Methods, De Gruyter, vol. 5(1), pages 79-101, January.

    More about this item

    NEP fields

    This paper has been announced in the following NEP Reports:

    Statistics

    Access and download statistics

    Corrections

    All material on this site has been provided by the respective publishers and authors. You can help correct errors and omissions. When requesting a correction, please mention this item's handle: RePEc:arx:papers:2409.00379. See general information about how to correct material in RePEc.

    If you have authored this item and are not yet registered with RePEc, we encourage you to do it here. This allows to link your profile to this item. It also allows you to accept potential citations to this item that we are uncertain about.

    If CitEc recognized a bibliographic reference but did not link an item in RePEc to it, you can help with this form .

    If you know of missing items citing this one, you can help us creating those links by adding the relevant references in the same way as above, for each refering item. If you are a registered author of this item, you may also want to check the "citations" tab in your RePEc Author Service profile, as there may be some citations waiting for confirmation.

    For technical questions regarding this item, or to correct its authors, title, abstract, bibliographic or download information, contact: arXiv administrators (email available below). General contact details of provider: http://arxiv.org/ .

    Please note that corrections may take a couple of weeks to filter through the various RePEc services.

    IDEAS is a RePEc service. RePEc uses bibliographic data supplied by the respective publishers.