Abstract
In practice, it is common that errors are correlated in the nonparametric regression model. Although many methods have been developed for addressing correlated errors, most of them rely on accurate estimation of correlation structure. A couple of methods have been proposed to avoid prior information of correlation structure to estimate regression function. However, the derivative estimation is also crucial to some practical applications. In this article, a bandwidth selection procedure is proposed for estimating both mean response and derivatives via kernel regression when correlated errors present. Both empirical support and theoretical justification are provided for the estimation procedure. Finally, we describe a Beijing temperature data example to illustrate the application of the proposed method.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
References
Altman NS (1990) Kernel smoothing of data with correlated errors. J Am Stat Assoc 85:749–759. https://doi.org/10.1080/01621459.1990.10474936
Carmack PS, Schucany WR, Spence JS, Gunst RF, Lin Q, Haley RW (2009) Far casting cross-validation. J Comput Graph Stat 18:879–893. https://doi.org/10.1198/jcgs.2009.07034
Carmack PS, Spence JS, Schucany WR (2012) Generalised correlated crossvalidation. J Nonparametric Stat 24:269–282. https://doi.org/10.1080/10485252.2012.655733
Charnigo R, Hall B, Srinivasan C (2011) A generalized Cp criterion for derivative estimation. Technometrics 53:238–253. https://doi.org/10.1198/TECH.2011.09147
Charnigo R, Srinivasan C (2011) Self-consistent estimation of mean response functions and their derivatives. Can J Stat 39:280–299. https://doi.org/10.1002/cjs.10104
Dai W, Tong T, Genton MG (2016) Optimal estimation of derivatives in nonparametric regression. J Mach Learn Res 17:5700–5724. https://doi.org/10.5555/2946645.3053446
De Brabanter K, Cao F, Gijbels I, Opsomer J (2018) Local polynomial regression with correlated errors in random design and unknown correlation structure. Biometrika 105:681–690. https://doi.org/10.1093/biomet/asy025
De Brabanter K, De Brabanter J, Suykens JA, De Moor B (2011) Kernel regression in the presence of correlated errors. J Mach Learn Res 12:1955–1976. https://doi.org/10.5555/1953048.2021062
Fan J, Gijbels I (1996) Local polynomial modelling and its applications: monographs on statistics and applied probability, vol 66. CRC Press, Boca Raton
Francisco-Fernandez M, Opsomer J, Vilar-Fernandez JM (2004) Plug-in bandwidth selector for local polynomial regression estimator with correlated errors. Nonparametric Stat 16:127–151. https://doi.org/10.1080/10485250310001622848
Gasser T, Müller H-G (1984) Estimating regression functions and their derivatives by the kernel method. Scand J Stat 11:171–185
Hall P, Van Keilegom I (2003) Using difference-based methods for inference in nonparametric regression with time series errors. J Royal Stat Soc Series B (Stat Methodol) 65:443–456. https://doi.org/10.1111/1467-9868.00395
Hart JD (1991) Kernel regression estimation with time series errors. J Royal Stat Soc Series B (Methodol) 53:173–187. https://doi.org/10.1111/j.2517-6161.1991.tb01816.x
Herrmann E, Gasser T, Kneip A (1992) Choice of bandwidth for kernel regression when residuals are correlated. Biometrika 79:783–795. https://doi.org/10.1093/biomet/79.4.783
Joo J-H, Qiu P (2009) Jump detection in a regression curve and its derivative. Technometrics 51:289–305. https://doi.org/10.1198/tech.2009.07163
Kim TY, Park BU, Moon MS, Kim C (2009) Using bimodal kernel for inference in nonparametric regression with correlated errors. J Multivar Anal 100:1487–1497. https://doi.org/10.1016/j.jmva.2008.12.012
Liu Y, De Brabanter K (2020) Smoothed nonparametric derivative estimation using weighted difference quotients. J Mach Learn Res 21:1–45
Loader C (2006) Local regression and likelihood. Springer Science & Business Media, Berlin
Mallows CL (1973) Some comments on Cp. Technometrics 15:661–675
Müller H-G (1984) Smooth optimum kernel estimators of densities, regression curves and modes. Ann Stat 12:766–774. https://doi.org/10.1214/aos/1176346523
Müller H-G, Stadtmüller U, Schmitt T (1987) Bandwidth choice and confidence intervals for derivatives of noisy data. Biometrika 74:743–749. https://doi.org/10.1093/biomet/74.4.743
Opsomer J, Wang Y, Yang Y (2001) Nonparametric regressin with correlated errors. Stat Sci 16:134–153. https://doi.org/10.1214/ss/1009213287
Park BU, Lee YK, Kim TY, Park C (2006) A simple estimator of error correlation in non-parametric regression models. Scand J Stat 33:451–462. https://doi.org/10.1111/j.1467-9469.2006.00506.x
Ramsay JO, Silverman BW (2002) Applied functional data analysis: methods and case studies, vol 77. Springer, Berlin
Simonoff JS (2012) Smoothing methods in statistics. Springer Science & Business Media, Berlin
Swain PS, Stevenson K, Leary A, Montano-Gutierrez LF, Clark IB, Vogel J, Pilizota T (2016) Inferring time derivatives including cell growth rates using gaussian processes. Nat Commun 7:1–8. https://doi.org/10.1038/ncomms13766
Zhang S, Guo B, Dong A, He J, Xu Z, Chen SX (2017) Cautionary tales on air-quality improvement in beijing. Proc Royal Soc A Math Phys Eng Sci 473:20170457. https://doi.org/10.1098/rspa.2017.0457
Acknowledgements
Sisheng Liu’s research is supported by the Research Foundation of Education Bureau of Hunan Province (Grant 22B0037). Jing Yang’s research is funded by the Natural Science Foundation of Hunan Province (Grant 2022JJ30368) and the National Natural Science Foundation of China (Grant 11801168).
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary Information
Below is the link to the electronic supplementary material.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Sisheng, L., Jing, Y. Kernel regression for estimating regression function and its derivatives with unknown error correlations. Metrika 87, 1–20 (2024). https://doi.org/10.1007/s00184-023-00901-9
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00184-023-00901-9