基于广义概念格的广义粗近似空间中规则的发现与提取

This paper proposes a new method of constructing generahzed concept lattice and producing rules from it in the generalized rough approximate space based on generalized similar relation which is more extensive than equivalent relation. Finally, a simple algorithm is presented to extract rules based on interesting measure.

利用广义粗近似挖掘缺省规则

由于对称性和传递性在一些情况下是不必要的,因此以仅有自反性的广义相似关系为基础研究缺省规则的发现是十分有意义的工作。本文首先给出新关系的形式和特点,再利用它得到挖掘缺省规则的广义粗近似框架,指出属性上的关系是反映客观世界之间联系的本质因素。