Time Series Analysis and Multivariate Statistical Analysis
Abstract
Research was carried out mainly in the areas of time series analysis and multivariate statistical analysis. The most important results in the first area apply to autoregressive and moving average processes. In the second area emphasis was on parameter consistency and elliptically contoured distributions. To estimate the parameters of the moving average model 16 different iterative procedures have been devised. These involve alternative parametrizations, time and frequency domain representations, Newton-Raphson and scoring approaches, and use of likelihoods and concentrated likelihoods. Properties of the likelihood function, as well as the estimates, have been derived. Keywords: Kalman filtering, Army research.
Document Details
- Document Type
- Technical Report
- Publication Date
- Nov 01, 1988
- Accession Number
- ADA202273
Entities
People
- Theodore W. Anderson
Organizations
- Stanford University