A Study on Developing the Acquisition Unit Cost Estimating Model of the Guided Weapon System |
Yonghyun Kim, Yongbok Lee, Wonil Jung, Dongkyu Kim, Sungjin Kang |
Korea National Defense University |
유도무기 획득단가 추정 모델 개발에 관한 연구 |
김용현, 이용복, 정원일, 김동규, 강성진 |
국방대학교 |
|
Abstract |
Cost estimates are necessary for government acquisition program to support decisions about funding, to develop annual budget requests and to validate resource requirements at key decision points. Many researches have been done about cost estimating technique recently. Parametric cost estimating models based on CERs(Cost Estimating Relationships) have been mainly used using regression method with historical data. However, there are many restrictions in developing Korean version CERs because the number of data points are too small. Specially, data collection and data management system are unstable in Korean defense environment, when developing CERs. In this research, we analyzed the historical data, and found some cost drivers in guided weapon system area. We developed the Acquisition Unit Cost CER using the regression to remove multicollinearity in the historical data. So we could overcome the restriction of the insufficient sample number. This research as a first attempt is meaningful in terms of obtaining our own Acquisition Unit Cost CER using historical cost and physical characteristic in Korean development environment. |
Key Words:
Regression Analysis, CER(Cost Estimation Relationship), Multicolinearity, Outlier, MRE, MMRE, PRED(${ell}$) |
|