Scholion.ai

Kernel Regression Estimation with Time Series Errors

Jeffrey D. Hart

Journal of the Royal Statistical Society: Series B (Statistical Methodology) · 1991 · Volume 53 · Issue 1

DOI: 10.1111/j.2517-6161.1991.tb01816.x

Abstract

Suppose that datay1, ...,ynare observed according to the model, wherefis a smooth function andis a sample from a zero-mean, covariance stationary time series. The problem of objectively choosing the bandwidth of a kernel estimate forfis addressed. It is shown both theoretically and by simulation that cross-validation produces extremely rough kernel estimates when the data are sufficiently positively correlated. This makes it inadvisable to use residuals from a cross-validated kernel estimate as a means of estimating the covariance function of the errors. Alternative methods of estimating the covariance function are proposed. In a simulation study, incorporating these estimated covariances into a risk estimation procedure leads to more efficient smoothing of positively correlated data.

Keywords

autoregressive process, bandwidth selection, mean average-squared error, mean-squared error, spectrum

Enable JavaScript for citation tools and additional paper interactions.