Forecast of the Second Destination Transportation Tonnage for the Air Force.
Abstract
This study analyzed the time series of second destination tonnage and suggests models to forecast them. An assortment of models were attempted including: exponential smoothing, polynominal regression, and Box-Jenkins, and the author found that a range of models can be used to forecast tonnage. The models were applied to eleven tonnage series, and two different types of forecasting models were recommended based on the principal of cost-effectiveness; single exponential smoothing for nine of the series, and decomposition of the remaining two. There is no implementation cost in terms of computer programming or training in following these recommendations. Only two values are stored for exponential smoothing--the current tonnage and the forecast for current period. Time series decomposition requires the use of seasonal indices only. Seasonal indices are provided by the author. (Author)
Document Details
- Document Type
- Technical Report
- Publication Date
- Feb 01, 1979
- Accession Number
- ADA107235
Entities
People
- Jack C. Hayya