Methods of Fitting Learning Curves to Lot Data Based on Assumptions and Techniques of Regression Analysis.

摘要

This thesis is primarily concerned with the formulae and mathematical procedures incorporated in the ICLOT and CALOT programs which were developed in 1967 to fit learning curves to historical lot data under the unit curve and cumulative average curve theories, respectively. The same formulae and procedures were also incorporated in the ICUNI and CAUNI programs, which were developed at the same time to fit learning curves to unit data. These programs have been used to support DCAA recommendations on cost estimates submitted by hundreds of defense contractors. This usage, plus the availability of these programs in the general libraries of services which account for most of the time-shared computer usage in the United States, has led to their adoption by many defense contractors and other time-shared computer customers.