AUTOMATIC LOCAL SMOOTHING FOR SPECTRAL DENSITY-ESTIMATION

This article uses local polynomial techniques to fit Whittle's likelih ood for spectral density estimation. Asymptotic sampling properties of the proposed estimators are derived, and adaptation of the proposed e stimator to the boundary effect is demonstrated. We show that the Whit tle likelihood-based estimator has advantages over the least-squares b ased log-periodogram. The bandwidth for the Whittle likelihood-based m ethod is chosen by a simple adjustment of a bandwidth selector propose d in Fan & Gijbels (1995). The effectiveness of the proposed procedure is demonstrated by a few simulated and real numerical examples. Our s imulation results support the asymptotic theory that the likelihood ba sed spectral density and log-spectral density estimators are the most appealing among their peers.