Accession Number : ADA288251

Title :   The Use of Time Series Forecasting in Contractor Performance Analysis.

Descriptive Note : Technical rept.,

Corporate Author : ARMY COMMUNICATIONS-ELECTRONICS COMMAND FORT MONMOUTH NJ

Personal Author(s) : D'Accardi, Richard J.

PDF Url : ADA288251

Report Date : NOV 1994

Pagination or Media Count : 22

Abstract : The application of time series forecasting is presented as a method of estimating either contract Estimates-at-Completion (EAC) or year-ahead contract costs based on information contained in monthly Contract Performance Reports (CPR). Considering the premise that the pattern of a contractor's expenditures becomes characteristic as the contract proceeds, a procedural approach is recommended for characterizing the cumulative Actual Cost of Work Performed (ACWP) using time series modeling. The recommended technique stresses the structuring of appropriate difference equations through the criterion of minimum residual variance. The main objectives of this report are: Development of a procedure for structuring forecasting models from non-stationary stochastic realizations. This specifically covers the identification and fitting of Autoregressive (AR), and integrated Autoregressive-Moving Averages (ARIMA) models; Illustration of the validity of the appropriate model through a sensitivity analysis using CPR information from two Communications-Electronics Command contracts.

Descriptors :   *CONTRACT ADMINISTRARION, *TIME SERIES ANALYSIS, MATHEMATICAL MODELS, PARAMETRIC ANALYSIS, COST ANALYSIS, PROBABILITY DISTRIBUTION FUNCTIONS, COMMUNICATION EQUIPMENT, CONTRACTORS, DIFFERENCE EQUATIONS, ELECTRONIC EQUIPMENT, FORECASTING, PERFORMANCE TESTS, AUTOCORRELATION, ANALYSIS OF VARIANCE.

Subject Categories : Administration and Management
      Economics and Cost Analysis
      Numerical Mathematics

Distribution Statement : APPROVED FOR PUBLIC RELEASE