Time Series Analysis
Abstract
This article addresses some of the concerns and questions about the use of time series analysis and covers some of the simpler, but useful techniques. To begin, time series analysis techniques are not deterministic techniques. Unlike regression, they do not reflect a causing relationship. Instead, they all take advantage of a pattern over time. In time series analysis, it is recognized that there are many causing variables. It we expect their magnitudes and effects to remain, we can accept that the net effect of all those variables can be summarized by the changing pattern of cost over time. We thus use time as a stand-in for all the causing variables. We believe that future costs can be predicted based on the pattern formed by past costs. When this sort of pattern over time has developed, and we believe that all the causing variables will continue to exert the same forces and move along their same paths, then time series analysis techniques provide an excellent alternative for the analyst.
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 01, 1982
- Accession Number
- ADP001242
Entities
People
- Roland D. Kankey
Organizations
- Air Force Institute of Technology