基于相似度的粗糙集近似算子快速求解

由于经典粗糙集只能处理精确分类问题,基于相似度的粗糙集模型被提出并用于解决不完备信息系统的相关问题.粗糙集通过近似算子对某一给定的概念进行近似表示,科学的求解这些算子对粗糙集理论的发展具有重要意义.本文提出一种新的近似算子快速求解方法,分析证明了所提快速方法比经典方法具有更高的求解效率.文章定义了元素覆盖度、集合覆盖度等概念,使用覆盖度等价关系可以将覆盖粗糙集转化为经典粗糙集,从而简化覆盖粗糙集的相关问题的解决.

Fast Solution of Similarity-based Rough Set Approximate Operators

Similarity-based rough set models are put forward to solve incomplete information systems because classical rough set is on-ly used to deal with precise classification. Rough set is used to approximately represent a certain concept by approximate operators, and getting these operators effectively is of great significance for the development of rough set thcory. A new fast solution of getting approximate operators is presented, and compared with classical method, the method proposed has higher efficiency. The covering degree of elements and sets are defined, and applying equivalent relation of covering degree can translate covering rough sets into clas-sical rough set easily, so that simplifying the theory of covering rough sets.