基于属性约简的群体评价信息集结模型研究

Group Evaluation Model of Information Aggregation Based on Attribute Reduction

针对群体评价问题中指标过多或指标相关性较高的情况,提出了基于粗糙集理论的指标约简法。根据指标属性的重要性大小确定指标的重要程度,约去冗余的、相关性高的评价指标,得到精简的指标体系,从而削减评价工作的工作量,提高评价工作的效率。对约简后的指标体系,运用基于到最优方案距离的评价方法进行综合评价。将所建立的综合评价模型应用到具体的群体评价问题中,并与其它方法结果进行比较,仿真结果验证了模型和方法的有效性与可行性,为群体评价问题提供了步骤上的创新。

For the issue of too many indicators or indicators with high correlation in group evaluation, a method of indicator reduction based on rough set is presented. The importance degrees of indicators are determined according to their importance of attributes, and the redundant indicators are reduced. Thus the workload of group evaluation is cut, and the work efficiency is improved. For the indicator system after reduction, comprehensive evaluation is car- ried out using the method of ranking distances from the ideal solution. Compared with the results of other methods, simulation resuhs show the feasibility and effectiveness of the method.