残缺数据型决策中的逆判方法及应用

Reverse judgement methods of fragmentary data and its application

研究基于残缺数据的排序决策中的逆判问题.在双因素方差分析和对数据进行可比化处理的基础上,结合样本信息改进传统特征根法确定专家权重的模型,进行方案实现分析.对比于传统排序方法,该方法考虑更多的因素影响,增强了决策科学性,增大了样本的分辨率,对专家的逆判结果更接近实际情况.为高缺失率的数据型决策中的逆判问题提供了一种有效的分析方法,但该算法的收敛性有待进一步改进.

The reverse judgement in order decision-making based on fragmentary data is discussed. Bifactor variance analysis and comparable treatment of the data are made. And then combined with the sample information to determine the weights of the judges by improving the traditional eigenvalue evaluation in order that the realization of project would be analyzed. The method considers the impact of more factors and enhances the scientific of deci- sion-making and increases the sample resolution compared to traditional methods. And the reverse judgement is closer to the actual situation. It has offered a kind of effective analysis method in high missing rate decision-making, but the convergence of this algorithm need to make a further discussion and study.