弱标记不完备决策系统的启发式增量约简算法  

Heuristic Increemental Attribute Reduction Algorithm for Weakly Labeled Incomplete Data

作　　者：郑颖春[1] 金莎 ZHENG Ying-chun;JIN Sha(School of Science,Xi’an University of Science and Technology,Xi’an 710600,China)

机构地区：[1]西安科技大学理学院,陕西西安710600

出　　处：《模糊系统与数学》2022年第2期121-130,共10页Fuzzy Systems and Mathematics

基　　金：国家自然科学基金资助项目(71473194);陕西省科技计划项目(2020CGXNG-013)。

摘　　要：针对实际决策系统中存在大量数据缺失、无类别标记,数据随时间动态变化的情形,提出属性与样本同时变化的增量式属性约简算法。将弱标记不完备决策系统分为有标记和无标记两个子系统,引入改进的属性区分关系定义,基于正域和容差关系概念,分别给出子系统中属性与样本动态变化时区分关系的增量式更新定理,并提出了弱标记不完备决策系统下的属性相对区分度增量式属性约简算法。通过实验分析表明本文算法的有效性和鲁棒性。There are often a lot of missing and unlabeled data in actual decision-making systems, and these data change dynamically over time. In response to this situation, an incremental attribute reduction algorithm is proposed in which attributes and samples change at the same time. The weakly labeled incomplete decision-making system is divided into two sub-systems, labeled and unlabeled, and an improved definition of attribute distinction relations is introduced. Based on the concept of positive domain and tolerance relationship, the incremental update methods of distinguishing the relationship between attributes and samples in the subsystem are given when dynamic changes are made. In addition, an incremental attribute reduction algorithm based on the relative discrimination of attributes in the weakly labeled incomplete decision-making system is proposed. Experimental analysis shows the effectiveness and completeness of the algorithm in this paper.

关 键 词：属性约简 区分关系 属性相对区分度 属性重要度 增量机制 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]