Electronic Journal of Statistics

Nonparametric conditional variance and error density estimation in regression models with dependent errors and predictors

Rafał Kulik and Cornelia Wichelhaus

Full-text: Open access

Abstract

This paper considers nonparametric regression models with long memory errors and predictors. Unlike in weak dependence situations, we show that the estimation of the conditional mean has influence on the estimation of both, the conditional variance and the error density. In particular, the estimation of the conditional mean has a negative effect on the asymptotic behaviour of the conditional variance estimator. On the other hand, surprisingly, estimation of the conditional mean may reduce convergence rates of the residual-based Parzen-Rosenblatt density estimator, as compared to the errors-based one. Our asymptotic results reveal small/large bandwidth dichotomous behaviour. In particular, we present a method which guarantees that a chosen bandwidth implies standard weakly dependent-type asymptotics. Our results are confirmed by an extensive simulation study. Furthermore, our theoretical lemmas may be used in different problems related to nonparametric regression with long memory, like cross-validation properties, bootstrap, goodness-of-fit or quadratic forms.

Article information

Source
Electron. J. Statist., Volume 5 (2011), 856-898.

Dates
First available in Project Euclid: 22 August 2011

Permanent link to this document
https://projecteuclid.org/euclid.ejs/1314018118

Digital Object Identifier
doi:10.1214/11-EJS629

Mathematical Reviews number (MathSciNet)
MR2831519

Zentralblatt MATH identifier
1274.62244

Subjects
Primary: 62G05: Estimation
Secondary: 62G08: Nonparametric regression 62E20: Asymptotic distribution theory 62M10: Time series, auto-correlation, regression, etc. [See also 91B84]

Keywords
long memory nonparametric regression random design conditional variance density estimation

Citation

Kulik, Rafał; Wichelhaus, Cornelia. Nonparametric conditional variance and error density estimation in regression models with dependent errors and predictors. Electron. J. Statist. 5 (2011), 856--898. doi:10.1214/11-EJS629. https://projecteuclid.org/euclid.ejs/1314018118


Export citation

References

  • [1] Florin Avram and Murad S. Taqqu. Noncentral limit theorems and Appell polynomials., Ann. Probab., 15(2):767–775, 1987.
  • [2] J. Beran and Y. Feng. Local polynomial estimation with a FARIMA-GARCH error process., Bernoulli, 7(5):733–750, 2001.
  • [3] T. Tony Cai and Lie Wang. Adaptive variance function estimation in heteroscedastic nonparametric regression., Ann. Statist., 36(5) :2025–2054, 2008.
  • [4] Ngai Hang Chan and Shiqing Ling. Residual empirical processes for long and short memory time series., Ann. Statist., 36(5) :2453–2470, 2008.
  • [5] Bing Cheng and P. M. Robinson. Density estimation in strongly dependent nonlinear time series., Statist. Sinica, 1(2):335–359, 1991.
  • [6] Fuxia Cheng. Asymptotic distributions of error density and distribution function estimators in nonparametric regression., J. Statist. Plann. Inference, 128(2):327–349, 2005.
  • [7] Sándor Csörgő and Jan Mielniczuk. Random-design regression under long-range dependent errors., Bernoulli, 5(2):209–224, 1999.
  • [8] Sándor Csörgő and Jan Mielniczuk. The smoothing dichotomy in random-design regression with long-memory errors based on moving averages., Statist. Sinica, 10(3):771–787, 2000.
  • [9] Sam Efromovich. Estimation of the density of regression errors., Ann. Statist., 33(5) :2194–2227, 2005.
  • [10] Jianqing Fan and Qiwei Yao. Efficient estimation of conditional variance functions in stochastic regression., Biometrika, 85(3):645–660, 1998.
  • [11] Jianqing Fan and Qiwei Yao., Nonlinear time series. Springer Series in Statistics. Springer-Verlag, New York, 2003. Nonparametric and parametric methods.
  • [12] C. Fraley, F. Leisch, and M. Maechler. fracdiff:, Fractionally differenced ARIMA aka ARFIMA(p,d,q) models, 2006. R package version 1.3-1.
  • [13] Jiti Gao and Qiying Wang. Long-range dependent time series specification., 2009.
  • [14] Liudas Giraitis and Donatas Surgailis. Central limit theorem for the empirical process of a linear sequence with long memory., J. Statist. Plann. Inference, 80(1-2):81–93, 1999.
  • [15] Hongwen Guo and Hira L. Koul. Asymptotic inference in some heteroscedastic regression models with long memory design and errors., Ann. Statist., 36(1):458–487, 2008.
  • [16] Peter Hall, Soumendra Nath Lahiri, and Jörg Polzehl. On bandwidth choice in nonparametric regression with both short- and long-range dependent errors., Ann. Statist., 23(6) :1921–1936, 1995.
  • [17] Peter Hall, Soumendra Nath Lahiri, and Young K. Truong. On bandwidth choice for density estimation with dependent data., Ann. Statist., 23(6) :2241–2263, 1995.
  • [18] Hwai-Chung Ho and Tailen Hsing. On the asymptotic expansion of the empirical process of long-memory moving averages., Ann. Statist., 24(3):992 –1024, 1996.
  • [19] Hwai-Chung Ho and Tailen Hsing. Limit theorems for functionals of moving averages., Ann. Probab., 25(4) :1636–1669, 1997.
  • [20] Hira Koul., Weighted Empirical and Residual Processes in Dynamical Nonlinear Models. Springer Series in Statistics. Springer-Verlag, New York, 2002.
  • [21] Hira L. Koul and Donatas Surgailis. Goodness-of-fit testing under long memory., J. Statist. Plann. Inference, 140(12) :3742–3753, 2010.
  • [22] Rafał Kulik. Nonparametric deconvolution problem for dependent sequences., Electron. J. Statist., 2:722–740, 2008.
  • [23] Rafal Kulik. Empirical process of long-range dependent sequences when parameters are estimated., J. Statist. Plann. Inference, 139(2):287–294, 2009.
  • [24] Rafal Kulik and Pawel Lorek. Empirical process of residuals for regression models with long memory errors., Preprint, 2010.
  • [25] Rafal Kulik and Pawel Lorek. Some results on random design regression with long memory errors and predictors., J. Statist. Plann. Inference, 141(1):508–523, 2011.
  • [26] Rafal Kulik and Cornelia Wichelhaus. Conditional variance estimation in regression models with long memory., 2010.
  • [27] S. N. Lahiri., Resampling methods for dependent data. Springer Series in Statistics. Springer-Verlag, New York, 2003.
  • [28] Han-Ying Liang and Si-Li Niu. Asymptotic properties of error density estimator in regression model under, α-mixing assumptions. Comm. Statist. Theory Methods, 38(6-7):761–783, 2009.
  • [29] Jan Mielniczuk and Wei Biao Wu. On random-design model with dependent errors., Statist. Sinica, 14(4) :1105–1126, 2004.
  • [30] Ursula U. Müller, Anton Schick, and Wolfgang Wefelmeyer. Estimating the error distribution function in semiparametric regression., Statist. Decisions, 25(1):1–18, 2007.
  • [31] Ursula U. Müller, Anton Schick, and Wolfgang Wefelmeyer. Estimating the error distribution function in nonparametric regression with multivariate covariates., Statist. Probab. Lett., 79(7):957–964, 2009.
  • [32] Natalie Neumeyer and Ingrid Van Keilegom. Estimating the error distribution in nonparametric multiple regression with applications to model testing., J. Multivariate Anal., 101(5) :1067–1078, 2010.
  • [33] D. Ruppert, S. J. Sheather, and M. P. Wand. An effective bandwidth selector for local least squares regression., J. Amer. Statist. Assoc., 90(432) :1257–1270, 1995.
  • [34] Murad S. Taqqu. Weak convergence to fractional Brownian motion and to the Rosenblatt process., Z. Wahrscheinlichkeitstheorie und Verw. Gebiete, 31:287–302, 1974/75.
  • [35] Lie Wang, Lawrence D. Brown, T. Tony Cai, and Michael Levine. Effect of mean on variance function estimation in nonparametric regression., Ann. Statist., 36(2):646–664, 2008.
  • [36] Wei Biao Wu and Jan Mielniczuk. Kernel density estimation for linear processes., Ann. Statist., 30(5) :1441–1459, 2002.
  • [37] Zhibiao Zhao and Wei Biao Wu. Confidence bands in nonparametric time series regression., Ann. Statist., 36(4) :1854–1878, 2008.