基于Rough集的数据库信息挖掘加权决策算法

Rough Set-based Weighted Decision Algorithm for Database Informa tion Mining

利用Rough集方法进行数据库中信息挖掘时,因为相同的记录表示的是同样的决策信息,可以将它消去,但由此得到的决策规则可能是片面的。文章提出了一种带有权重的决策算法,该方法利用记录出现的频率作为权重系数,导出了带有全重的决策算法,该方法计算简便、实用,可提高最终决策算法的可靠性。

When mining database for information by rough set method,since the same records make same contribution to decision,generally only one of them remains while others can be cancelled out ,which in a sense causes the loss of information,and the resulting decision rules may be one-sided.In this paper a weighted decision algorithm is present ,which makes use of frequency of the record as weight.This algorithm is computationally simple and efficiently enhances the decision reliability.