The Annals of Statistics

Spline Smoothing and Optimal Rates of Convergence in Nonparametric Regression Models

Paul Speckman

Full-text: Open access

Abstract

Linear estimation is considered in nonparametric regression models of the form $Y_i = f(x_i) + \varepsilon_i, x_i \in (a, b)$, where the zero mean errors are uncorrelated with common variance $\sigma^2$ and the response function $f$ is assumed only to have a bounded square integrable $q$th derivative. The linear estimator which minimizes the maximum mean squared error summed over the observation points is derived, and the exact minimax rate of convergence is obtained. For practical problems where bounds on $\|f^{(q)}\|^2$ and $\sigma^2$ may be unknown, generalized cross-validation is shown to give an adaptive estimator which achieves the minimax optimal rate under the additional assumption of normality.

Article information

Source
Ann. Statist. Volume 13, Number 3 (1985), 970-983.

Dates
First available: 12 April 2007

Permanent link to this document
http://projecteuclid.org/euclid.aos/1176349650

JSTOR
links.jstor.org

Digital Object Identifier
doi:10.1214/aos/1176349650

Mathematical Reviews number (MathSciNet)
MR803752

Zentralblatt MATH identifier
0585.62074

Subjects
Primary: 62J05: Linear regression
Secondary: 62G35: Robustness 41A15: Spline approximation

Keywords
Cross-validation mean square linear estimation nonparametric regression splines

Citation

Speckman, Paul. Spline Smoothing and Optimal Rates of Convergence in Nonparametric Regression Models. The Annals of Statistics 13 (1985), no. 3, 970--983. doi:10.1214/aos/1176349650. http://projecteuclid.org/euclid.aos/1176349650.


Export citation