An Application of Linear Regression to the Allocation of General Purpose Vehicles
Abstract
The purpose of this thesis was to develop a mathematical model which could assist in the allocation of general purpose vehicles. The linear regression model developed was used, through the AFIT CSC computer system, within the SAS statistical software program. Data was obtained from two AFSC bases. Their input, through the SAS interaction, led to the following model: Y (predicted maintenance cost) = beat sub 0 + beta sub 1 X sub 1 + beat sub 2 x sub 2 where X sub 1 equals age and X sub 2 equals average miles driven. The analysis demonstrated the statistical significance of the model. It also highlighted the potential for its application to include all bases within a command. The data limitation, only two bases in the sample, restricted the analysis from making any macro statements; however, it appears from the minor differences in regression coefficients, as well as similar mean fleet ages and mean miles traveled, the current distribution system is an effective one. This is to say that the current system of allocation used by the AFSC/LGT, the fair share method is effective at keeping maintenance costs, as well as vehicle assets, balanced across the AFSC fleet.
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 01, 1986
- Accession Number
- ADA174566
Entities
People
- John H. Golden
Organizations
- Air Force Institute of Technology