摘要
针对传统评估指标筛选方法适用性差、主观性强,并且需要大量测试数据等问题,提出了粗糙集属性约简和灰色关联聚类相结合的指标筛选方法。利用灰色关联聚类解决了经典粗糙集理论中的等价关系分类方法仅仅适用于离散数据的问题,并针对灰色聚类临界值难以确定的问题,采用统计量法解决,两种方法的优势互补实现了评估指标的有效筛选。最后结合作战指挥系统的评估指标体系筛选案例,对方法的合理性和有效性进行了验证。
For the problem of traditional index screening methods are poor applicability and strong subjectivity, as well as a large number of test data, a new index screening method of the combination of rough set and grey relational cluster theory are proposed. Using grey cluster theory instead of equivalence relations of classical rough set classification method solved the question that the classical attribute reduction of rough set only applies to the problem of discrete data. The statistics method is used to solve the question that the grey clustering critical value is difficult to be determined, and this method are complementary to achieve the effective screening of the evaluation index. Finally, though the evaluation index system of integrated command system as an example, the rationality and validity of the method are verified.
出处
《火力与指挥控制》
CSCD
北大核心
2018年第1期37-42,共6页
Fire Control & Command Control
基金
国家"863"计划重点基金资助项目(2012AA7010213)
关键词
粗糙集
灰色关联度
灰色聚类临界值
统计量
rough set, grey relational, grey clustering critical, statistics
