A Geometric Unification of Distributionally Robust Covariance Estimators: Shrinking the Spectrum by Inflating the Ambiguity Set

References

• Atkinson C , Mitchell AF (1981) Rao’s distance measure. Sankhyā: Indian J. Statist. Series A 43(3):345–365.Google Scholar
• Blanchet J , Murthy K , Nguyen VA (2021a) Statistical analysis of Wasserstein distributionally robust estimators. Carlsson JG, ed. Emerging Optimization Methods and Modeling Techniques with Applications (INFORMS, Cantonsville, MD), 227–254.Link, Google Scholar
• Blanchet J , Murthy K , Si N (2021b) Confidence regions in Wasserstein distributionally robust estimation. Biometrika 109(2):295–315.Crossref, Google Scholar
• Bodnar T , Gupta AK , Parolya N (2016) Direct shrinkage estimation of large dimensional precision matrix. J. Multivariate Anal. 146:223–236.Crossref, Google Scholar
• Bui N , Nguyen D , Yue MC , Nguyen VA (2025) Coverage-validity-aware algorithmic recourse. Oper. Res. 73(6):3294–3310.Link, Google Scholar
• Dennis JE Jr , Schnabel RB (1996) Numerical Methods for Unconstrained Optimization and Nonlinear Equations (SIAM, Philadelphia).Crossref, Google Scholar
• Donoho D , Gavish M , Johnstone I (2018) Optimal shrinkage of eigenvalues in the spiked covariance model. Ann. Statist. 46(4):1742–1778.Crossref, Google Scholar
• Eguiluz VM , Chialvo DR , Cecchi GA , Baliki M , Apkarian AV (2005) Scale-free brain functional networks. Phys. Rev. Lett. 94(1):018102.Crossref, Google Scholar
• Gao R (2023) Finite-sample guarantees for Wasserstein distributionally robust optimization: Breaking the curse of dimensionality. Oper. Res. 71(6):2291–2306.Link, Google Scholar
• Gao R , Chen X , Kleywegt AJ (2024) Wasserstein distributionally robust optimization and variation regularization. Oper. Res. 72(3):1177–1191.Link, Google Scholar
• Ghaoui LE , Oks M , Oustry F (2003) Worst-case value-at-risk and robust portfolio optimization: A conic programming approach. Oper. Res. 51(4):543–556.Link, Google Scholar
• Givens CR , Shortt RM (1984) A class of Wasserstein metrics for probability distributions. Michigan Math. J. 31(2):231–240.Crossref, Google Scholar
• Hardy GH , Littlewood JE , Pólya G (1952) Inequalities (Cambridge University Press, London).Google Scholar
• Hastie T , Tibshirani R , Friedman JH (2009) The Elements of Statistical Learning (Springer, New York).Crossref, Google Scholar
• Hoerl AE , Kennard RW (1970) Ridge regression: Biased estimation for nonorthogonal problems. Technometrics 12(1):55–67.Crossref, Google Scholar
• Jagannathan R , Ma T (2003) Risk reduction in large portfolios: Why imposing the wrong constraints helps. J. Finance 58(4):1651–1683.Crossref, Google Scholar
• James W , Stein C (1992) Estimation with quadratic loss. Kotz S, Johnson NL, eds. Breakthroughs in Statistics (Springer, New York), 443–460.Crossref, Google Scholar
• Jeffreys H (1946) An invariant form for the prior probability in estimation problems. Proc. Roy. Soc. London A 186(1007):453–461.Crossref, Google Scholar
• Kalman RE (1960) A new approach to linear filtering and prediction problems. J. Basic Engrg. 82(1):35–45.Crossref, Google Scholar
• Kuhn D , Mohajerin Esfahani P , Nguyen VA , Shafieezadeh-Abadeh S (2019) Wasserstein distributionally robust optimization: Theory and applications in machine learning. Netessine S, ed. Operations Research & Management Science in the Age of Analytics (INFORMS, Cantonsville, MD), 130–166.Link, Google Scholar
• Kullback S (1997) Information Theory and Statistics (Courier Corporation, Gloucester, MA).Google Scholar
• Ledoit O , Wolf M (2003) Improved estimation of the covariance matrix of stock returns with an application to portfolio selection. J. Empirical Finance 10(5):603–621.Crossref, Google Scholar
• Ledoit O , Wolf M (2004a) Honey, I shrunk the sample covariance matrix. J. Portfolio Management 30(4):110–119.Crossref, Google Scholar
• Ledoit O , Wolf M (2004b) A well-conditioned estimator for large-dimensional covariance matrices. J. Multivariate Anal. 88(2):365–411.Crossref, Google Scholar
• Ledoit O , Wolf M (2012) Nonlinear shrinkage estimation of large-dimensional covariance matrices. Ann. Statist. 40(2):1024–1060.Crossref, Google Scholar
• Ledoit O , Wolf M (2017) Nonlinear shrinkage of the covariance matrix for portfolio selection: Markowitz meets Goldilocks. Rev. Financial Stud. 30(12):4349–4388.Crossref, Google Scholar
• Ledoit O , Wolf M (2020) Analytical nonlinear shrinkage of large-dimensional covariance matrices. Ann. Statist. 48(5):3043–3065.Crossref, Google Scholar
• Ledoit O , Wolf M (2022) Quadratic shrinkage for large covariance matrices. Bernoulli 28(3):1519–1547.Crossref, Google Scholar
• Lohweg V (2013) Banknote authentication dataset. UCI machine learning repository. https://doi.org/10.24432/C55P57.Google Scholar
• Mantegna RN (1999) Hierarchical structure in financial markets. Eur. Phys. J. B 11:193–197.Crossref, Google Scholar
• Markowitz H (1952) Portfolio selection. J. Finance 7(1):77–91.Google Scholar
• Mohajerin Esfahani P , Kuhn D (2018) Data-driven distributionally robust optimization using the Wasserstein metric: Performance guarantees and tractable reformulations. Math. Programming 171(1–2):115–166.Crossref, Google Scholar
• Nguyen VA , Kuhn D , Mohajerin Esfahani P (2022) Distributionally robust inverse covariance estimation: The Wasserstein shrinkage estimator. Oper. Res. 70(1):490–515.Link, Google Scholar
• Nguyen VA , Shafieezadeh-Abadeh S , Filipović D , Kuhn D (2021) Mean-covariance robust risk measurement. Preprint, submitted December 18, https://arxiv.org/abs/2112.09959.Google Scholar
• Nguyen VA , Shafieezadeh-Abadeh S , Yue MC , Kuhn D , Wiesemann W (2019a) Calculating optimistic likelihoods using (geodesically) convex optimization. Advances in Neural Information Processing Systems (Curran Associates Inc., Red Hook, NY), 13920–13931.Google Scholar
• Nguyen VA , Shafieezadeh Abadeh S , Yue MC , Kuhn D , Wiesemann W (2019b) Optimistic distributionally robust optimization for nonparametric likelihood approximation. Advances in Neural Information Processing Systems (Curran Associates Inc., Red Hook, NY), 13942–13953.Google Scholar
• Pearson K (1895) Note on regression and inheritance in the case of two parents. Proc. Roy. Soc. London 58(347–352):240–242.Crossref, Google Scholar
• Perlman M (2007) STAT 542: Multivariate statistical analysis. Lecture Notes, University of Washington, Seattle.Google Scholar
• Rajaratnam B , Vincenzi D (2016) A theoretical study of Stein’s covariance estimator. Biometrika 103(3):653–666.Crossref, Google Scholar
• Rockafellar R (1997) Convex Analysis (Princeton University Press, Princeton, NJ).Google Scholar
• Shafieezadeh-Abadeh S , Kuhn D , Mohajerin Esfahani P (2019) Regularization via mass transportation. J. Machine Learn. Res. 20(103):1–68.Google Scholar
• Shafieezadeh-Abadeh S , Aolaritei L , Dörfler F , Kuhn D (2023) Nash equilibria, regularization and computation in optimal transport-based distributionally robust optimization. Preprint, submitted March 7, https://arxiv.org/abs/2303.03900.Google Scholar
• Shafieezadeh-Abadeh S , Nguyen VA , Kuhn D , Mohajerin Esfahani P (2018) Wasserstein distributionally robust Kalman filtering. Advances in Neural Information Processing Systems (Curran Associates Inc., Red Hook, NY), 8483–8492.Google Scholar
• Sharpe WF (1963) A simplified model for portfolio analysis. Management Sci. 9(2):277–293.Link, Google Scholar
• Stein C (1975) Estimation of a covariance matrix, Rietz Lecture. Proc. 39th Annual Meeting Institute Math. Statist. (Institute of Mathematical Statistics).Google Scholar
• Taşkesen B , Iancu D , Koçyiğit Ç , Kuhn D (2023) Distributionally robust linear quadratic control. Advances in Neural Information Processing Systems (Curran Associates Inc., Red Hook, NY), 18613–18632.Google Scholar
• Taskesen B , Yue MC , Blanchet J , Kuhn D , Nguyen VA (2021) Sequential domain adaptation by synthesizing distributionally robust experts. Meila M, Zhang T, eds. Proc. Internat. Conf. Machine Learn . (PMLR, New York), 10162–10172.Google Scholar
• Taylor R (1990) Interpretation of the correlation coefficient: A basic review. J. Diagnostic Medical Sonography 6(1):35–39.Crossref, Google Scholar
• Touloumis A (2015) Nonparametric Stein-type shrinkage covariance matrix estimators in high-dimensional settings. Comput. Statist. Data Anal. (Oxford) 83:251–261.Crossref, Google Scholar
• van der Vaart HR (1961) On certain characteristics of the distribution of the latent roots of a symmetric random matrix under general conditions. Ann. Math. Statist. 32(3):864–873.Crossref, Google Scholar
• van Wieringen WN (2015) Lecture notes on ridge regression. Preprint, submitted September 30, https://arxiv.org/abs/1509.09169.Google Scholar
• Villani C (2008) Optimal Transport: Old and New (Springer, Berlin).Google Scholar
• Vu H , Tran T , Yue MC , Nguyen VA (2022) Distributionally robust fair principal components via geodesic descents. Proc. Internat. Conf. Learn. Representations (OpenReview.net).Google Scholar
• Wolberg W , Mangasarian O , Street N , Street W (1992) Breast cancer Wisconsin (diagnostic) dataset. UCI machine learning repository. https://doi.org/10.24432/C5DW2B.Google Scholar
• Zorzi M (2017) Robust Kalman filtering under model perturbations. IEEE Trans. Automated Control 62(6):2902–2907.Crossref, Google Scholar