[关键词]
[摘要]
信用评价周期包括指标考核周期和总的信用评价周期。结合长周期工程特点,基于神经网络模型(LVQ)对信用评价指标的时间敏感性分类以及对优化周期组合进行了系统研究,以期获得信用评价成本最小下的周期确定方法。研究结果表明：基于指标的时间敏感性分类,对不同时间敏感性的指标实行不同的指标考核周期。在指标分类周期组合( 1、2、3、6 )时,能够实现每个信用评价成本最小化,最佳的信用评价周期为T=6。
[Key word]
[Abstract]
The credit evaluation cycle includes the index assessment cycle and the total credit evaluation cycle. Based on the characteristics of long period engineering, the time sensitivity classification of credit evaluation indexes and the optimal cycle combination are systematically studied based on the neural network model (LVQ), in order to obtain the periodic determination method under the minimum cost of credit evaluation. The research results show that: Based on the time sensitive classification of indicators, different index assessment periods are implemented for different time sensitive indicators. When the index classification cycle combination (1, 2, 3, 6) can achieve the minimum cost of each credit evaluation, the best credit evaluation cycle is T=6.
[中图分类号]
C931；F426.92
[基金项目]
国家自然科学基金项目 “建设工程项目交易模式创新路径与增值机理研究”（71402045）；国务院南水北调办公室委托项目“南水北调建筑物工程质量监管与通水期运行管理研究”(20155018702)
