Date of Award


Document Type


Degree Name

Master of Science in Cost Analysis


Department of Systems Engineering and Management

First Advisor

Brandon M. Lucas, PhD.


This thesis examines an approach to characterizing various expenditure profiles for the Air Force Installation Contracting Agency’s Operations and Maintenance Appropriated Funds. Using naive, seasonal naive, trailing moving average, exponential smoothing, linear regression, and autoregressive integrated moving average (ARIMA) forecasting methods, the paper evaluates multiple error measures over one fiscal year to find the most precise model for each level of analysis. Levels of analysis included the Air Force enterprise and level 1 category levels, as well as an illustrative approach to Information and Technology spend at the level 2 subcategory, major command, and base levels. Optimal model characteristics were used to compare expenditure profile patterns at the different levels. In general, the more a unit can customize its algorithms, the more accurately it can capture its respective expenditure profile. The more localized the level of spend, the less applicable the aggregate models become, and different sub-groups have more personalized patterns.

AFIT Designator


DTIC Accession Number