Volume 2 · Issue 4 (2025)
A Data Driven Selecting Rule for the Bandwidth of Regression Estimation with Generalized Additive Noise
Jiahao Mou, Cong Wu
School of Science, Hubei university of technology, Wuhan, 430068, China
Abstract: For the problem of bandwidth selection in nonparametric regression, most of the existing methods carried out theoretical analysis and numerical calculation with a fixed bandwidth. However, the selection of an appropriate bandwidth and the guarantee of good performance depend heavily on the parameters of the smoothness of regression function, which are difficult to calculate in practice. To overcome this problem, a novel and efficient data-driven selecting rule is proposed to adaptively determine the appropriate bandwidth. It turns out that the bandwidth only loses the lnn factor in terms of convergence rate by selecting rule.
Keywords: Data-driven; Regression estimation; The selecting rule of bandwidth
References
[1]
Alexander, M.: Deconvolution Problems in Nonparametric Statistics. Springer, Berlin (2009). https://doi.org/10.1007/978-3-540-87557-4
[2]
Guo, H., Liu, Y.: Convergence rates of multivariate regression estimators with errors-in-variables. NUMERICAL FUNCTIONAL ANALYSIS AND OPTIMIZATION. 38, 1564–1588 (2017). https://doi.org/10.1080/01630563.2017.1349145
[3]
Bouzebda, S., El-hadjali, T.: Uniform convergence rate of the kernel regression estimator adaptive to intrinsic dimension in presence of censored data. JOURNAL OF NONPARAMETRIC STATISTICS. 32, 864–914 (2020). https://doi.org/10.1080/10485252.2020.1834107
[4]
Pinelis, I.: Monotonicity preservation properties of kernel regression estimators. STATISTICS & PROBABILITY LETTERS. 177 (2021) https://doi.org/10.1016/j.spl.2021.109157
[5]
Ali, T.H.: Modification of the adaptive nadaraya-watson kernel method for nonparametric regression (simulation study). COMMUNICATIONS IN STATISTICS-SIMULATION AND COMPUTATION. 51, 391–403 (2022) https://doi.org/10.1080/03610918.2019.1652319
[6]
Liu, Y., Wu, C.: Point-wise wavelet estimation in the convolution structure density model. JOURNAL OF FOURIER ANALYSIS AND APPLICATIONS. 26 (2020) https://doi.org/10.1007/s00041-020-09794-y
[7]
Kong, D., Zhao, S., Wang, W.: Nonparametric derivative estimation with bimodalkernels under correlated errors. COMPUTATIONAL STATISTICS. 39, 1847– 1865 (2024) https://doi.org/10.1007/s00180-023-01419-4
[8]
Chen, J., Kou, J.: Pointwise estimation of anisotropic regression functions using wavelets with data-driven selection rule. MATHEMATICS. 12
[9]
Kim, H.-J., Chen, H.-S., Midthune, D., Wheeler, B., Buckman, D.W., Green, D., Byrne, J., Luo, J., Feuer, E.J.: Data-driven choice of a model selection method in joinpoint regression. JOURNAL OF APPLIED STATISTICS. 50, 1992–2013 (2023) https://doi.org/10.1080/02664763.2022.2063265
[10]
Li, P., Jiao, Y., Lu, X., Kang, L.: A data-driven line search rule for support recovery in high-dimensional data analysis. COMPUTATIONAL STATISTICS & DATA ANALYSIS. 174 (2022) https://doi.org/10.1016/j.csda.2022.107524
[11]
Liu, X., Song, Y., Cheng, H.-F., Zhang, K.: A bootstrap-based bandwidth selection rule for kernel quantile estimators. COMPUTATIONAL STATISTICS. 40, 4037–4058 (2025) https://doi.org/10.1007/s00180-024-01582-2
[12]
Liu, X., Song, Y., Zhang, K.: An exact bootstrap-based bandwidth selection rule for kernel quantile estimators. COMMUNICATIONS IN STATISTICS-SIMULATION AND COMPUTATION. 53, 3699–3720 (2024) https://doi.org/10.1080/03610918.2022.2110595
[13]
Bagkavos, D., Patil, P.N., Ramanathan, T.V.: Local linear estimation and plug-in bandwidth selection for the ratio of two hazard rate functions. JOURNAL OF NONPARAMETRIC STATISTICS. (2025) https://doi.org/10.1080/10485252.2025.2466650
[14]
El-Dakkak, O., Feng, S., Wahbah, M., EL-Fouly, T.H.M., Zahawi, B.: Combinatorial method for bandwidth selection in wind speed kernel density estimation. IET RENEWABLE POWER GENERATION. 13, 1670–1680 (2019) https://doi.org/10.1049/iet-rpg.2018.5643
[15]
Wylupek, G.: Data-driven k-sample tests. TECHNOMETRICS. 52, 107–123 (2010) https://doi.org/10.1198/TECH.2009.08101
Download PDF