Distance-covariance-based tests for heteroscedasticity in nonlinear regressions

We use distance covariance to introduce novel consistent tests of heteroscedasticity for nonlinear regression models in multidimensional spaces. The proposed tests require no user-defined regularization, which are simple to implement based on only pairwise distances between points in the sample and are applicable even if we have non-normal errors and many covariates in the regression model. We establish the asymptotic distributions of the proposed test statistics under the null and alternative hypotheses and a sequence of local alternatives converging to the null at the fastest possible parametric rate. In particular, we focus on whether and how the estimation of the finite-dimensional unknown parameter vector in regression functions will affect the distribution theory. It turns out that the asymptotic null distributions of the suggested test statistics depend on the data generating process, and then a bootstrap scheme and its validity are considered. Simulation studies demonstrate the versatility of our tests in comparison with the score test, the Cramer-von Mises test, the Kolmogorov-Smirnov test and the Zheng-type test. We also use the ultrasonic reference block data set from National Institute for Standards and Technology of USA to illustrate the practicability of our proposals.

[1]  Robert P. Sherman,et al.  Maximal Inequalities for Degenerate $U$-Processes with Applications to Optimization Estimators , 1994 .

[2]  Daojiang He,et al.  Omnibus Model Checks of Linear Assumptions through Distance Covariance , 2021 .

[3]  David S. Matteson,et al.  Independent Component Analysis via Distance Covariance , 2013, 1306.4911.

[4]  C. Hsiao,et al.  A CONSISTENT TEST FOR CONDITIONAL HETEROSKEDASTICITY IN TIME-SERIES REGRESSION MODELS , 2001, Econometric Theory.

[5]  Lixing Zhu,et al.  Pairwise distance-based heteroscedasticity test for regressions , 2020, Science China Mathematics.

[6]  Xiaofeng Shao,et al.  Martingale Difference Divergence Matrix and Its Application to Dimension Reduction for Stationary Multivariate Time Series , 2018 .

[7]  Xianyang Zhang,et al.  Distance Metrics for Measuring Joint Dependence with Application to Causal Inference , 2017, Journal of the American Statistical Association.

[8]  Runze Li,et al.  Projection correlation between two random vectors , 2017, Biometrika.

[9]  S. Weisberg,et al.  Diagnostics for heteroscedasticity in regression , 1983 .

[10]  Lixing Zhu,et al.  TESTING HETEROSCEDASTICITY FOR REGRESSION MODELS BASED ON PROJECTIONS , 2018, Statistica Sinica.

[11]  Xiaoming Huo,et al.  Fast Computing for Distance Covariance , 2014, Technometrics.

[12]  Lixing Zhu,et al.  Heteroscedasticity checks for regression models , 2001 .

[13]  A. Ullah,et al.  A NONPARAMETRIC GOODNESS-OF-FIT-BASED TEST FOR CONDITIONAL HETEROSKEDASTICITY , 2013, Econometric Theory.

[14]  Bodhisattva Sen,et al.  Testing independence and goodness-of-fit in linear models , 2013 .

[15]  D. Pollard,et al.  Simulation and the Asymptotics of Optimization Estimators , 1989 .

[16]  Xu Zheng.  Testing heteroscedasticity in nonlinear and nonparametric regressions , 2009 .

[17]  Bernhard Schölkopf,et al.  Measuring Statistical Dependence with Hilbert-Schmidt Norms , 2005, ALT.

[18]  Maria L. Rizzo,et al.  Partial Distance Correlation with Methods for Dissimilarities , 2013, 1310.2926.

[19]  X. Shao,et al.  Testing mutual independence in high dimension via distance covariance , 2016, 1609.09380.

[20]  Assessing the Adequacy of Variance Function in Heteroscedastic Regression Models , 2007, Biometrics.

[21]  R. Jennrich.  Asymptotic Properties of Non-Linear Least Squares Estimators , 1969 .

[22]  Xiangrong Yin,et al.  Sufficient Dimension Reduction via Distance Covariance , 2016 .

[23]  Keith Knight,et al.  Limiting distributions for $L\sb 1$ regression estimators under general conditions , 1998 .

[24]  Lixing Zhu,et al.  Model diagnosis for parametric regression in high-dimensional spaces , 2008 .

[25]  Heping Zhang,et al.  Ball Covariance: A Generic Measure of Dependence in Banach Space , 2019, Journal of the American Statistical Association.

[26]  Heping Zhang,et al.  Conditional Distance Correlation , 2015, Journal of the American Statistical Association.

[27]  Maria L. Rizzo,et al.  Measuring and testing dependence by correlation of distances , 2007, 0803.4101.

[28]  Anil K. Ghosh,et al.  A distribution-free two-sample run test applicable to high-dimensional data , 2014 .

[29]  Winfried Stute,et al.  Nonparametric model checks for regression , 1997 .

[30]  C. Priebe,et al.  Network dependence testing via diffusion maps and distance-based correlations , 2017, Biometrika.

[31]  Ing Rj Ser.  Approximation Theorems of Mathematical Statistics , 1980 .

[32]  J. Escanciano.  A CONSISTENT DIAGNOSTIC TEST FOR REGRESSION MODELS USING PROJECTIONS , 2006, Econometric Theory.

[33]  Maria L. Rizzo,et al.  Brownian distance covariance , 2009, 1010.0297.

[34]  P. Sen.  Estimates of the Regression Coefficient Based on Kendall's Tau , 1968 .

[35]  Konstantinos Fokianos,et al.  Testing independence for multivariate time series via the auto‐distance correlation matrix , 2018 .

[36]  Runze Li,et al.  Feature Screening via Distance Correlation Learning , 2012, Journal of the American Statistical Association.

[37]  Carey E. Priebe,et al.  From Distance Correlation to Multiscale Graph Correlation , 2017, Journal of the American Statistical Association.

[38]  Xiaofeng Shao,et al.  Martingale Difference Correlation and Its Use in High-Dimensional Variable Screening , 2014 .

[39]  Cédric Heuchenne,et al.  Nonlinear Regression With Censored Data , 2007, Technometrics.

[40]  Ingrid Keilegom,et al.  Goodness-of-fit tests in parametric regression based on the estimation of the error distribution , 2008 .

[41]  J. Zheng,et al.  A consistent test of functional form via nonparametric estimation techniques , 1996 .

[42]  B. Sen,et al.  On a nonparametric notion of residual and its applications , 2014, 1409.3886.

[43]  Anne Leucht,et al.  Consistency of general bootstrap methods for degenerate U-type and V-type statistics , 2009, J. Multivar. Anal..