Minimax-Optimal Policy Learning Under Unobserved Confounding

References

• Aronow P, Lee D (2012) Interval estimation of population means under unknown but bounded probabilities of sample selection. Biometrika 100(1):235–240.Google Scholar
• Athey S, Tibshirani J, Wager S (2019) Generalized random forests. Ann. Statist. 47(2):1148–1178.Crossref, Google Scholar
• Bakour SH, Williamson J (2015) Latest evidence on using hormone replacement therapy in the menopause. Obstetrician Gynaecologist 17(1):20–28.Crossref, Google Scholar
• Bartlett PL, Bousquet O, Mendelson S (2005) Local Rademacher complexities. Ann. Statist. 33(4):1497–1537.Crossref, Google Scholar
• Bertsimas D, Dunn J (2017) Optimal classification trees. Machine Learn. 106:1039–1082.Crossref, Google Scholar
• Bertsimas D, Kallus N, Weinstein AM, Zhuo YD (2016) Personalized diabetes management using electronic medical records. Diabetes Care 40(2):210–217.Google Scholar
• Beygelzimer A, Langford J (2009) The offset tree for learning with partial labels. Proc. 15th ACM SIGKDD Internat. Conf. Knowledge Discovery Data Mining. (ACM, New York), 129–138.Google Scholar
• Blake T, Nosko C, Tadelis S (2015) Consumer heterogeneity and paid search effectiveness: A large-scale field experiment. Econometrica 83(1):155–174.Crossref, Google Scholar
• Brookhart MA, Sturmer T, Glynn RJ, Rassen J, Schneeweiss S (2010) Confounding control in healthcare database research: Challenges and potential approaches. Medical Care 48(6 Suppl):S114–S120.Crossref, Google Scholar
• Charnes A, Cooper W (1962) Programming with linear fractional functionals. Naval Res. Logist. Quart. 9(3–4):181–186.Crossref, Google Scholar
• Chernozhukov V, Chetverikov D, Demirer M, Duflo E, Hansen C (2018) Double machine learning for treatment and causal parameters. Econometrics J. 21(1):C1–C68.Google Scholar
• Dudik M, Erhan D, Langford J, Li L (2014) Doubly robust policy evaluation and optimization. Statist. Sci. 29(4):485–511.Crossref, Google Scholar
• Dudley R (1987) Universal Donsker classes and metric entropy. Ann. Probab. 15(4):1306–1326.Google Scholar
• Fogarty C, Hasegawa R (2019) An extended sensitivity analysis for heterogeneous unmeasured confounding. Ann. Appl. Statist. 13(2):767–796.Crossref, Google Scholar
• Fogarty CB, Small DS (2016) Sensitivity analysis for multiple comparisons in matched observational studies through quadratically constrained linear programming. J. Amer. Statist. Assoc. 111(516):1820–1830.Crossref, Google Scholar
• Giné E, Nickl R (2016) Mathematical Foundations of Infinite-Dimensional Statistical Models, vol. 40 (Cambridge University Press, Cambridge, UK).Crossref, Google Scholar
• Golea M, Bartlett PL, Lee WS, Mason L (1998) Generalization in decision trees and DNF: Does size matter? Jordan M, Kearns M, Solla S, eds. Adv. Neural Inform. Processing Systems (MIT Press, Boston), 259–265.Google Scholar
• Gordon BR, Zettelmeyer F, Bhargava N, Chapsky D (2019) A comparison of approaches to advertising measurement: Evidence from big field experiments at Facebook. Marketing Sci. 38(2):193–225.Link, Google Scholar
• Hájek J (1971) Comment on “An essay on the logical foundations of survey sampling, part one.” Godambe VP, Thompson, ME, eds. The Foundations of Survey Sampling, vol. 236 (Holt, Rinehart and Winston, Toronto).Google Scholar
• Hasegawa R, Small D (2017) Sensitivity analysis for matched pair analysis of binary data: From worst case to average case analysis. Biometrics 73(4):1424–1432.Crossref, Google Scholar
• Hirano K, Porter JR (2019) Statistical decision rules in econometrics. Durlauf S, Hansen L, Heckman J, Matzkin R, eds. Handbook of Econometrics (Elsevier, Amsterdam), 7.Google Scholar
• Ho T-H, Lim N, Reza S, Xia X (2017) OM forum—Causal inference models in operations management. Manufacturing Service Oper. Management 19(4):509–525.Link, Google Scholar
• Hoffman MA, Williams MS (2011) Electronic medical records and personalized medicine. Human Genetics 130(1):33–39.Crossref, Google Scholar
• Horvitz D, Thompson D (1952) A generalization of sampling without replacement from a finite universe. J. Amer. Statist. Assoc. 47(260):663–685.Crossref, Google Scholar
• Hsu JY, Small DS (2013) Calibrating sensitivity analyses to observed covariates in observational studies. Biometrics 69(4):803–811.Crossref, Google Scholar
• Imbens G, Rubin D (2015) Causal Inference for Statistics, Social, and Biomedical Sciences (Cambridge University Press, Cambridge, England).Crossref, Google Scholar
• Kallus N (2017a) Recursive partitioning for personalization using observational data. Precup D, Teh YW, eds. Proc. 34th Internat. Conf. Machine Learn, vol. 70 (PMLR), 1789–1798.Google Scholar
• Kallus N (2017b) Balanced policy evaluation and learning. Adv. Neural Inform. Processing Systems, 8895–8906.Google Scholar
• Kallus N, Zhou A (2018a) Confounding-robust policy improvement. Adv. Neural Inform. Processing Systems (PMLR), 9269–9279.Google Scholar
• Kallus N, Zhou A (2018b) Policy evaluation and optimization with continuous treatments. Internat. Conf. Artificial Intelligence Statist. 1243–1251.Google Scholar
• Kitagawa T, Tetenov A (2018) Who should be treated? Empirical welfare maximization methods for treatment choice. Econometrica 86(2):591–616.Crossref, Google Scholar
• Künzel S, Sekhon J, Bickel P, Yu B (2019) Metalearners for estimating heterogeneous treatment effects using machine learning. Proc. Natl. Acad. Sci. USA 116(10):4156–4165.Google Scholar
• Lawlor DA, Smith GD, Ebrahim S (2004) Commentary: The hormone replacement-coronary heart disease conundrum: Is this the death of observational epidemiology? Internat. J. Epidemiology 33(3):464–467.Crossref, Google Scholar
• Li L, Chu W, Langford J, Wang X (2011) Unbiased offline evaluation of contextual-bandit-based news article recommendation algorithms. Proc. Fourth ACM Internat. Conf. Web Search Data Mining (ACM, New York), 297–306.Google Scholar
• Manski C (2005) Social Choice with Partial Knowledge of Treatment Response (The Econometric Institute Lectures, Princeton, NJ).Google Scholar
• Manski C (2008) Identification for Prediction and Decision (Harvard University Press, Cambridge, MA).Crossref, Google Scholar
• Manson JE, Chlebowski RT, Stefanick ML, Aragaki AK, Rossouw JE, Prentice RL, Anderson G, et al.. (2013) The women’s health initiative hormone therapy trials: Update and overview of health outcomes during the intervention and post-stopping phases. JAMA 310(13):1353–1368.Google Scholar
• Masten M, Poirier A (2018) Identification of treatment effects under conditional partial independence. Econometrica 86(1):317–351.Crossref, Google Scholar
• Miratrix LW, Wager S, Zubizarreta JR (2018) Shape-constrained partial identification of a population mean under unknown probabilities of sample selection. Biometrika 105(1):103–114.Crossref, Google Scholar
• Nie X, Wager S (2017) Learning objectives for treatment effect estimation. Preprint, submitted 2017, https://arxiv.org/pdf/1712.04912.pdf.Google Scholar
• Pedersen AT, Ottesen B (2003) Issues to debate on the women’s health initiative (WHI) study. Epidemiology or randomized clinical trials—Time out for hormone replacement therapy studies? Human Reproduction 18(11):2241–2244.Crossref, Google Scholar
• Petrik M, Ghavamzadeh M, Chow Y (2016) Safe policy improvement by minimizing robust baseline regret. 29th Conf. Neural Inform. Processing Systems (PMLR), 2298–2306.Google Scholar
• Prentice RL, Pettinger M, Anderson GL (2005) Statistical issues arising in the women’s health initiative. Biometrics 61(4):899–911.Crossref, Google Scholar
• Qian M, Murphy SA (2011) Performance guarantees for individualized treatment rules. Ann. Statist. 39(2):1180–1210.Crossref, Google Scholar
• Robins JM, Rotnitzky A, Zhao LP (1994) Estimation of regression coefficients when some regressors are not always observed. J. Amer. Statist. Assoc. 89(427):846–866.Crossref, Google Scholar
• Rosenbaum P (2002) Observational Studies. Springer Series Statistics (Springer, New York).Google Scholar
• Rossouw JE, Manson JE, Kaunitz AM, Anderson GL (2013) Lessons learned from the Women’s Health Initiative trials of menopausal hormone therapy. Obstetrics Gynecology 121(1):172–176.Crossref, Google Scholar
• Rubin D (1974) Estimating causal effect of treatments in randomized and nonrandomized studies. J. Ed. Psych. 66(5):688–701.Crossref, Google Scholar
• Rubin DB (1980) Comments on “Randomization analysis of experimental data: The Fisher randomization test comment.” J. Amer. Statist. Assoc. 75(371):591–593.Google Scholar
• Shalit U, Johansson FD, Sontag D (2017) Estimating individual treatment effect: Generalization bounds and algorithms. Internat. Conf. Machine Learning (PMLR), 3076–3085.Google Scholar
• Stoye J (2009) Minimax regret treatment choice with finite samples. J. Econometrics 151(1):70–81.Crossref, Google Scholar
• Stoye J (2012) Minimax regret treatment choice with limited validity of experiments or with covariates. J. Econometrics 166(1):138–156.Crossref, Google Scholar
• Swaminathan A, Joachims T (2015a) Batch learning from logged bandit feedback through counterfactual risk minimization. J. Machine Learn. Res. 16(52):1731–1755.Google Scholar
• Swaminathan A, Joachims T (2015b) The self-normalized estimator for counterfactual learning. Adv. Neural Inform. Processing Systems (PMLR), 3231–3239.Google Scholar
• Tan Z (2012) A distributional approach for causal inference using propensity scores. J. Amer. Statist. Assoc. 101(476):1619–1637.Google Scholar
• Thomas P, Theocharous G, Ghavamzadeh M (2015) High confidence policy improvement. Proc. 32nd Internat. Conf. Machine Learn (PMLR), 2380–2388.Google Scholar
• Ustun B, Rudin C (2015) Supersparse linear integer models for optimized medical scoring systems. Machine Learn. 102(3):349–91.Crossref, Google Scholar
• Van Der Vaart AW, Wellner JA (1996) Weak convergence. Weak Convergence and Empirical Processes (Springer, New York), 16–28.Crossref, Google Scholar
• Wager S, Athey S (2017a) Efficient policy learning. Preprint, submitted 2017, https://arxiv.org/abs/1702.02896.Google Scholar
• Wager S, Athey S (2017b) Estimation and inference of heterogeneous treatment effects using random forests. J. Amer. Statist. Assoc. 113(523)1228–1242.Crossref, Google Scholar
• Wang YX, Agarwal A, Dudik M (2017) Optimal and adaptive off-policy evaluation in contextual bandits. Proc. Neural Inform. Processing System (PMLR), 3589–3597.Google Scholar
• Zhao Q, Small DS, Bhattacharya BB (2019) Sensitivity analysis for inverse probability weighting estimators via the percentile bootstrap. J. Roy. Statist. Soc. Ser. B. Statist. Methodology 81(4):735–761.Crossref, Google Scholar