Local asymptotics for polynomial spline regression
Open Access
- 30 September 2003
- journal article
- research article
- Published by Institute of Mathematical Statistics in The Annals of Statistics
- Vol. 31 (5), 1600-1635
- https://doi.org/10.1214/aos/1065705120
Abstract
In this paper we develop a general theory of local asymptotics for least squares estimates over polynomial spline spaces in a regression problem. The polynomial spline spaces we consider include univariate splines, tensor product splines, and bivariate or multivariate splines on triangulations. We establish asymptotic normality of the estimate and study the magnitude of the bias due to spline approximation. The asymptotic normality holds uniformly over the points where the regression function is to be estimated and uniformly over a broad class of design densities, error distributions and regression functions. The bias is controlled by the minimum L-infinity norm of the error when the target regression function is approximated by a function in the polynomial spline space that is used to define the estimate. The control of bias relies on the stability in L-infinity norm of L-2 projections onto polynomial spline spaces. Asymptotic normality of least squares estimates over polynomial or trigonometric polynomial spaces is also treated by the general theory. In addition, a preliminary analysis of additive models is provided.Keywords
This publication has 32 references indexed in Scilit:
- The L2 Rate of Convergence for Event History Regression with Time‐dependent CovariatesScandinavian Journal of Statistics, 1998
- Triogram ModelsJournal of the American Statistical Association, 1998
- The Use of Polynomial Splines and Their Tensor Products in Multivariate Function EstimationThe Annals of Statistics, 1994
- Interaction Spline Models and Their Convergence RatesThe Annals of Statistics, 1991
- Asymptotics for Doubly Flexible Logspline Response ModelsThe Annals of Statistics, 1991
- Large-Sample Inference for Log-Spline ModelsThe Annals of Statistics, 1990
- Residual Variance and Residual Pattern in Nonlinear RegressionBiometrika, 1986
- The Dimensionality Reduction Principle for Generalized Additive ModelsThe Annals of Statistics, 1986
- A Bound on the L ∞ -Norm of L 2 -Approximation by Splines in Terms of a Global Mesh RatioMathematics of Computation, 1976
- Optimal L ∞ Error Estimates for Galerkin Approximations to Solutions of Two-Point Boundary Value ProblemsMathematics of Computation, 1975