摘要
从大量数据发现知识时 ,属性约简是一个关键问题 .本文提出了一种新的算法用于粗集中的属性约简 .该算法基于 Hu的差别矩阵 ,在对差别矩阵进行化简的基础上 ,先得到核 ,然后在逐步减小的差别矩阵中挑选出现最频繁的属性加入直到成为一个约简 .文中还对求核的正确性、算法的正确性进行了证明 ,同时对算法复杂度进行了分析 。
Knowledge reduction is an important issue when dealing with huge amounts of data. This paper introduces a new algorithm for reduction of attribute. Based on the modified discernibility matrix due to Hu, we first get the core attributes and then add the most frequent attribute in the changing discernibility matrix gradually until we find a reduct. we also give proofs for calculating of core and the whole algorithm, and analyze the complexity of the algorithm. At last, an example is given to show the validity of the algorithm.
出处
《小型微型计算机系统》
CSCD
北大核心
2003年第3期523-526,共4页
Journal of Chinese Computer Systems
