基于证据理论的综合评价定性指标测度优化研究

综合评价定性指标测度以评价者偏好为重要依据。围绕群体综合评价定性指标测度中,个体评价者偏好测度不充分与实践指导性不足以及群体偏好集成的个体权重难以确定等问题,提出一种基于证据理论的定性指标测度优化方法。首先,构建基于证据理论的偏好信息分布测度思路,利用以有效子集为基础的半可加离散分布,统一测度评价者各类型偏好选择及其信度。其次,利用迭代方法从重要性和可靠性两个方面,求解基于主客观信息的个体权重,并提出一种相适应的群体分布偏好集成方法。上述方法通过提高个体偏好信息测度与集成的充分性及实践指导性,实现优化综合评价定性指标测度的目标。最后,通过实例验证方法的有效性。

A Group Evaluation Method of Qualitative Indicator Measurement Based on Evidence Theory

In the face of qualitative indicator measurement in comprehensive evaluation, a method of expert group evaluation based on Evidence Theory is proposed in this paper.Firstly, a probability inference frame with qualitative indicator measurement is constructed to analyze all kinds of subjective information of a single expert based on different acknowledge.Secondly, a model of subjective information measurement is established and then an iteration method is used to acquire expert weights based on subjective and objective information.Thirdly, a new method of group information aggregation is developed to obtain a qualitative indicator measurement result of an interval probability distribution form by the whole group.Finally, an example shows the validity for proposed method.