不完全信息下基于交叉评价的灰色关联决策模型

Grey relational analysis models with incomplete information based on cross-evaluation

借鉴DEA交叉评价思想,构建不完全信息下的灰色关联决策模型.决策者对决策对象无"偏好"时,假设各决策对象均具有使"自身评价值最大化,竞争对手评价值最小化"的"自利性"特点;决策者对决策对象有"偏好"时,假设各决策对象均具有使"自身客观偏好与决策者的主观偏好的接近程度最小化"的一致目标.各决策对象按照最有利于自身的权重对所有决策对象进行交叉评价,集结后的交叉评价结果作为决策基准.新模型能公平体现各决策对象的"发言权"和竞争优势,决策结果易于被决策对象和决策者同时接受.

Learning from cross-evaluation method in DEA,the grey relational analysis models with incomplete information are established.If evaluators have no subjective preference,assumes each of the evaluated objects is self-serving to maximize its evaluation value and minimize the value of its every rival. If evaluators have subjective preference,supposes that all evaluated objects have consistent goal minimize the deviation between its objective value and subjective preferences of evaluators.Each of evaluated objects cross-evaluated others adopting weights which could compete for its goal.Integrated values of cross-evaluation results are used as decision criteria.The new models could reflect the＂voice＂and competitive advantage of evaluated objects fairly,and are easy to be accepted by evaluated objects and evaluators at the same time.