用DEA优化主成分回归的寿命周期费用建模  

Modeling for life cycle cost by using DEA to optimizing principal components regression

作　　者：张小海[1] 金家善[1] 耿俊豹[1] 陈国卫[1] 孙林凯[1] 张军[1] 

机构地区：[1]海军工程大学船舶与动力学院,武汉430033

出　　处：《海军工程大学学报》2011年第1期32-36,83,共6页Journal of Naval University of Engineering

基　　金：中国博士后科研基金资助项目(20080431380)

摘　　要：回归统计建模方法是寿命周期费用建模中常用的方法,回归统计建模方法本质上是对数据平均趋势的估算,无法回避"依据错误的数据得到错误的模型"的根本问题。为此,提出用DEA方法对数据进行评价,剔除无效数据,将有效的数据用于主成分回归的建模方法。该方法能有效克服干扰数据对提取成分的影响,弥补主成分分析的不足。通过实例计算并与PCR、PLSR进行了比较分析,结果表明:DEA主成分回归建模精度高于PCR和PLSR。The regression statistical modeling,one of the modeling methods of the life cycle cost（LCC）,is used for estimating the average trend of the data essentially,which can not avoid the pro-blem about obtaining the wrong model according to the error data.Therefore,the improved principal component regression（PCR） by means of data envelopment analysis（DEA） was proposed to overcome this problem,which could evaluate the efficiency of the data and remove the inefficiency data for PCR modeling.The new approach can overcome the effect of the disturbed data in filtering principal components in the independent variables.Compared with the method of PCR and partial least squares regression（PLSR）,the improved PCR with DEA is much higher in precision.

关 键 词：寿命周期费用 数据包络分析 主成分回归 偏最小二乘回归 

分 类 号：N945.11[自然科学总论—系统科学]