Official websites use .gov
A .gov website belongs to an official government organization in the United States.
Secure .gov websites use HTTPS
A lock (
) or https:// means you’ve safely connected to the .gov website. Share sensitive information only on official, secure websites.
Topics
Data & Maps
Surveys & Programs
Resource Library
A General Analysis of Watson's Minimax Procedure for Component Model Selection in Non-Stationary ARMA Model-Based Seasonal Adjustment
A General Analysis of Watson's Minimax Procedure for Component Model Selection in Non-Stationary ARMA Model-Based Seasonal Adjustment
RR84-33
Introduction
In applications of the Hillmer-Tiao model-based signal extraction approach to seasonal adjustment (see Hillmer, Bell and Tiao (1983) or Burma; (1980)), there usually is a white noise component which must be divided among the seasonal and nonseasonal components, or assigned wholly to one of them (see the example in section 4). This assignment determines the covariance structure which must be specified before the optimal estimates, or, essentially equivalently, the filters used to obtain them, can be determined. Recently, Watson (1984) proposed the use of a minimax criterion related to mean square component estimation-error for making this assignment. His approach is quite attractive, due in part to additional appealing properties he demonstrates for some of the solutions to his minimax problem.
Others in Series
Working Paper
Maximizing the Overlap Between Surveys When Information is Incomplete
Maximizing the Overlap Between Surveys When Information is Incomplete
Working Paper
A Note on the Equivalence of Models with Monthly Means and Models w...
A Note on the Equivalence of Models with Monthly Means and Models with (0,1,1)12 Seasonal Parts with Θ = 1
Working Paper
A Study of Pre-Adjusted Transformations
A Study of Pre-Adjusted Transformations
Related Information
WORKING PAPER
Statistical Research Reports and Studies
Page Last Revised - October 28, 2021
✕
Is this page helpful?
Yes
No
Yes
No✕
NO THANKS
255 characters maximum
255 characters maximum reached
255 characters maximum reached
✕
Thank you for your feedback.
Comments or suggestions?
Comments or suggestions?
