Accession Number : ADA187127

Title :   A Comparison of Fitting Techniques for the Cumulative Average Learning Curves with Historical Lot Cost Data.

Descriptive Note : Master's thesis,

Corporate Author : AIR FORCE INST OF TECH WRIGHT-PATTERSON AFB OH SCHOOL OF SYSTEMS AND LOGISTICS

Personal Author(s) : Jones, John K

PDF Url : ADA187127

Report Date : Sep 1987

Pagination or Media Count : 214

Abstract : The technique used to fit cost data to the cumulative average learning curve can have an impact on the accuracy of the estimates provided. This research tested two specific fitting techniques, the elementary technique (ET) and the Calot technique, in an attempt to determine which technique provides the greater accuracy when used to fit historical lot cost data to the cumulative average learning curve. Both techniques were evaluated on there ability to fit total lot costs and predict last lot costs. The ET and Calot fitting techniques were both used to fit the historical lot costs for 66 systems to the cumulative average learning curve. A comparison of the two techniques indicates that Calot estimates total lot costs with greater accuracy more frequently and with a significantly lower standard error then the ET technique. Calot also demonstrated the ability to estimate the last lot costs more accurately more frequently than 50% of the time. (Theses)

Descriptors :   *LEARNING CURVES, *CURVE FITTING, ACCURACY, COMPARISON, COSTS, ERRORS, ESTIMATES, THESES, ALGORITHMS, FORMULATIONS, AIR FORCE PERSONNEL

Subject Categories : Numerical Mathematics

Distribution Statement : APPROVED FOR PUBLIC RELEASE