A Self-Consistent Approach for Nonparametric Spectral Density Estimation with Missing Data

• Activity Number: 578
• Type: Topic Contributed
• Date/Time: Wednesday, August 4, 2010 : 2:00 PM to 3:50 PM
• Sponsor: Section on Nonparametric Statistics
• Author(s): Zhengyuan Zhu*+ and Thomas C.M. Lee
• Companies: Iowa State University and University of California, Davis
• Keywords: missing data ; periodogram smoothing ; self-consistency ; nonparametric spectrum estimation

Abstract

Spectral density estimation is an important problem that arises in many different application areas. Data collected from these applications often contain missing values, which forbids the use of many powerful periodogram-based nonparametric spectral analysis techniques. In this paper we apply the self-consistency principle to develop a new method for nonparametric spectrum estimation with missing data which retain the maximum amount of information in the data. This method can be coupled with any complete data nonparametric spectrum estimation procedure, including kernel smoothing, wavelet and spline estimators. A simulation study and application to a sediment data are used to illustrate the practical performance of the method.