CONSISTENCY OF CROSS-VALIDATION WHEN THE DATA ARE CURVES Academic Article

abstract

• 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. 1993.

authors

• Hart, Jeffrey
• Wehrly, Thomas

published proceedings

• STOCHASTIC PROCESSES AND THEIR APPLICATIONS

author list (cited authors)

• HART, J. D., & WEHRLY, T. E.

citation count

• 29

complete list of authors

• HART, JD||WEHRLY, TE

publication date

• April 1993

publisher

• Elsevier  Publisher

keywords

• Bandwidth Selection
• Covariance Function
• Gaussian Processes
• Kernel Estimator

Digital Object Identifier (DOI)

• 10.1016/0304-4149(93)90080-N

start page

• 351

end page

• 361

volume

• 45

issue

• 2

URL

• http://dx.doi.org/10.1016/0304-4149(93)90080-n