R–fuzzy隶属度粗糙近似集的优势测度理论
Advantage measure theory of R-fuzzy rough membership set
摘要点击 145  全文点击 165  投稿时间:2017-10-16  修订日期:2018-04-23
查看全文  查看/发表评论  下载PDF阅读器
DOI编号  10.7641/CTA.2018.70746
  2019,36(2):286-294
中文关键词  R-fuzzy集  粗糙集  优势测度  智能感知  模式分类
英文关键词  R-fuzzy sets  Rough sets  Advantage measure  Perceptual recognition  Pattern visualization
基金项目  国家自然科学基金项目(61374043, 61603392);江苏省自然科学基金项目(BK20150199,BK20160275)
学科分类代码  
作者单位E-mail
李守军 中国矿业大学 lishoujunbox@126.com 
马小平 中国矿业大学  
杨春雨 中国矿业大学  
中文摘要
      给出了一种R-fuzzy集隶属度的权重比较与量化方法, 提出了优势测度概念, 解决了粗糙近似集中隶属度的重要性难以确定的问题. 首先给出优势测度的定义, 然后, 研究了优势测度的性质, 指出了优势测度1型模糊集的本质属性. 优势测度不仅实现了R-fuzzy集隶属度的量化, 而且成为R-fuzzy集与2型模糊集联系的纽带. 通过隶属度的优势测度, 实现了人类感知领域中的群体共识与个性认识的区分, 反过来通过优势测度的可视化, 可以对人类不同感知下的隶属度值给出合理的推断与比较. 最后, 通过声音感知实验研究给出了优势测度可视化的特点及操作方法, 讨论了不同职业测试组对于同一声音的理解在隶属度数值上的差异. 对于涉及人类感知与模式辨识的应用领域具有较易的操作性与较强的实用性.
英文摘要
      To solve the problem how to determine the importance of rough approximate membership, the concept of advantage measure is presented by bringing forward the weight comparison and quantitative method of R-fuzzy set membership degree. Firstly, the definition of the advantage measure is given, then, the properties of the advantage measure are investigated, and its essential attribute of type-1 fuzzy sets is pointed out. Advantage measure has not only realized the quantification of membership degree of R-fuzzy sets, but also becomes the tie between R-fuzzy sets and 2 type fuzzy sets. Through the advantage measure of membership degree, the distinction between group consensus and personality recognition in the field of human perception is realized. In turn, through the visualization of the advantage measure, reasonable inference and comparison to the degree of membership of different human perception can be given. Finally, the characteristics and operation methods of the advantage measure visualization are given by the sound sensing experiment. The differences in the membership degree of different profession groups for the same loudness were discussed, showing its operable and practicable characteristics in the application of human perception and pattern recognition.