A COMPARATIVE STUDY OF PREDICTION TECHNIQUES,
Abstract
Methods for improving the prediction of demand, specifically demand for aircraft spare parts, are sought where the prediction is based on an estimate of the underlying pattern of demands per program element. Seven prediction techniques were compared: (1) division of the total number of demands by number of flying hours in the experience period; (2) the demands are used to enter a table giving the upper 90% confidence level for the mean of a Poisson variable; (3) a nonparametric Poisson technique using as a constant the value of the normal deviate for the 90% confidence interval; (4) a technique using an equation similar to that used in (3); (5) an adaptation of the minimum Normit Chi-Square method (J. Berkson, 'Tables for use in estimating the normal distribution function by Normit analysis,' Biometrika, 44:411 (1957)); (6) a technique similar to (5) but using the transformation log (1 + age); and (7) the actuarial method of the Air Force for predicting engine failures (T. O. 00-25-128 of April 18, 1958).
Document Details
- Document Type
- Technical Report
- Publication Date
- Apr 29, 1960
- Accession Number
- AD0616659
Entities
People
- Bernice B. Brown
Organizations
- RAND Corporation