CONSISTENCY OF PARAMETER-ESTIMATES IN A LINEAR TIME-SERIES MODEL
Abstract
The basic conditions under which consistency in a time series analysis can be assured are stated. Some narrower conditions are derived from these, which may be easier to verify in some applications. These conditions show the existence of consistent least squares estimates also for some exponentially increasing regression vectors. Applications are made to polymonial, exponential, and trigonometric regression. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Oct 01, 1960
- Accession Number
- AD0248658
Entities
People
- Friedhelm Eicker
Organizations
- University of North Carolina at Chapel Hill