基于α优势关系的多粒度建模及对象排序方法  

Multi-Granularity Modeling and Object Ranking Based on α-Dominance Relationship

作　　者：张晓燕[1] 雷凉 徐伟华 ZHAGN Xiaoyan;LEI Liang;XU Weihua(College of Artificial Intelligent,Southwest University,Chongqing 400715,China)

机构地区：[1]西南大学人工智能学院,重庆400715

出　　处：《海南热带海洋学院学报》2023年第2期73-83,共11页Journal of Hainan Tropical Ocean University

基　　金：国家自然科学基金项目(61976245)。

摘　　要：为了探讨在多粒度视角下区间值序信息系统中数据分类和样本排序的问题,在区间值序信息系统中建立了α优势关系,研究了该优势关系导出的区间值信息系统多粒度粗糙集,进一步建立了排序预测模型,探索了其重要性质,给出了区间值数据分类的新方法。利用该预测模型,可解决用区间值序信息系统对新增对象计算排序得分时产生过于复杂的问题。最后,用实验分析验证了模型的可行性。In order to probe into the problem of data classification and sample ranking in the interval-valued order information system from the perspective of multi-granularity,a new method was proposed based on α-dominance relationship.The multi-granularity rough set,which was derived from this dominance relationship in interval-valued information system,was investigated.Moreover,the ranking prediction model was established,its important properties were explored,and a new method of interval-valued data classification was given for the system.The complex problem of the object ranking for newly added objects can be tackled by this prediction model in the interval value ordered information system.Finally,the feasibility of the model was verified by experimental analysis.

关 键 词：区间值序信息系统 多粒度粗糙集 对象排序 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]