CONSISTENCY OF CROSS-VALIDATION WHEN THE DATA ARE CURVES

被引：30

作者：

HART, JD

WEHRLY, TE

机构：

[1] Department of Statistics, Texas A and M University, College Station, TX

来源：

STOCHASTIC PROCESSES AND THEIR APPLICATIONS | 1993年 / 45卷 / 02期

关键词：

KERNEL ESTIMATOR; GAUSSIAN PROCESSES; COVARIANCE FUNCTION; BANDWIDTH SELECTION;

D O I：

10.1016/0304-4149(93)90080-N

中图分类号：

O21 [概率论与数理统计]; C8 [统计学];

学科分类号：

020208 ; 070103 ; 0714 ;

摘要：

Suppose one observes a random sample of n continuous time Gaussian processes on the interval [0, 1]; in other words, each observation is a curve. Of interest is estimating the common mean function of the processes by a kernel smoother. The bandwidth of the kernel estimator is chosen by a version of cross-validation in which deleting an observation means deleting one of the n curves. It is shown that using this form of cross-validation leads to an asymptotically optimal choice of bandwidth. This result is contrasted with the inconsistency of cross-validation in a seemingly more tractable problem.

引用

页码：351 / 361

页数：11

共 15 条

[1]

EUBANK RL, 1992, ANN STATIST

[2]

HALL P, 1992, J ROY STATIST SOC B

[3] HOW FAR ARE AUTOMATICALLY CHOSEN REGRESSION SMOOTHING PARAMETERS FROM THEIR OPTIMUM [J].