Local Influence Detection of Conditional Mean Dependence

Abstract

This article is focused on the problem to measure and test the conditional mean dependence of a response variable on a predictor variable. A local influence detection approach is developed combining with the martingale difference divergence (MDD) metric, and an efficient wild bootstrap implementation is given. The obtained new metric of the conditional mean dependence holds the merits of MDD, while it is more sensitive than the original one, and leads to a powerful test to nonlinear relationships. It is shown by simulations that the proposed test can achieve higher power for general conditional mean dependence relationships even in high-dimensional settings. Theoretical asymptotic properties of the local influence test statistic are given, and a real data analysis is also presented for further illustration. The localization idea could be combined with other conditional mean dependence metrics.

Keywords

Conditional mean independence / Martingale difference divergence / Local influence / Nonlinear dependence / 62G10 / 62G20

References

[1] Arcones MA, Giné E. Limit theorems for U-processes. Ann. Probab., 1993, 21(3): 1494-1542

[2] Fang, L., Yuan, Q., Ye, C., Yin, X.: High-dimensional variable screening via conditional martingale difference divergence. arXiv preprint arXiv:2206.11944 (2022)

[3] Free S, O’Higgins P, Maudgil D, Dryden I, Lemieux L, Fish D, Shorvon S. Landmark-based morphometrics of the normal adult brain using mri. Neuroimage, 2001, 13(5): 801-813

[4] Kendall DG. Shape manifolds, procrustean metrics, and complex projective spaces. Bull. Lond. Math. Soc., 1984, 16(2): 81-121

[5] Kosorok MR. Introduction to empirical processes and semiparametric inference, 2007, New York, Springer

[6] Lai T, Zhang Z, Wang Y. A kernel-based measure for conditional mean dependence. Comput. Stat. Data Anal., 2021, 160: 107246

[7] Lee C, Zhang X, Shao X. Testing conditional mean independence for functional data. Biometrika, 2020, 107(2): 331-346

[8] Li R, Xu K, Zhou Y, Zhu L. Testing the effects of high-dimensional covariates via aggregating cumulative covariances. J. Am. Stat. Assoc., 2022

[9] Lyons R. Distance covariance in metric spaces. Ann. Probab., 2013, 41(53284-3305

[10] Nolan D, Pollard D. Functional limit theorems for U-processes. Ann. Probab., 1988, 16(3): 1291-1298

[11] Pan W, Wang X, Zhang H, Zhu H, Zhu J. Ball covariance: a generic measure of dependence in banach space. J. Am. Stat. Assoc., 2020, 115(529): 307-317

[12] Park T, Shao X, Yao S. Partial martingale difference correlation. Electron. J. Stat., 2015, 9(1): 1492-1517

[13] Shao X, Zhang J. Martingale difference correlation and its use in high-dimensional variable screening. J. Am. Stat. Assoc., 2014, 109(507): 1302-1318

[14] Székely GJ, Rizzo ML, Bakirov NKB. Measuring and testing dependence by correlation of distances. Ann. Stat., 2007, 35(6): 2769-2794

[15] Vaart AWVD, Wellner JA. Weak convergence and empirical processes, 1996, New York, Springer

[16] Victor H, Victor H, la Peña D, de la Peña V, Giné E. Decoupling: from dependence to independence, 1999, New York, Springer

[17] Wang G, Zhu K, Shao X. Testing for the martingale difference hypothesis in multivariate time series models. J. Bus. Econ. Stat., 2021, 40: 1-15

[18] Zhang X, Yao S, Shao X. Conditional mean and quantile dependence testing in high dimension. Ann. Stat., 2018, 46(1): 219-246

[19] Zhou T, Zhu L, Xu C, Li R. Model-free forward screening via cumulative divergence. J. Am. Stat. Assoc., 2020, 115(531): 1393-1405

[20] Zhu J, Pan W, Zheng W, Wang X. Ball: an r package for detecting distribution difference and association in metric spaces. J. Stat. Softw., 2021, 97(1): 1-31

Funding

National natrural science foundation of China(12271014)

RIGHTS & PERMISSIONS

School of Mathematical Sciences, University of Science and Technology of China and Springer-Verlag GmbH Germany, part of Springer Nature