摘要
属性约简是知识发现中的关键问题之一。为了有效地获取最小相对约简,该文基于Hu的区分矩阵,即以属性核为起点,通过向属性核不断添加重要程度最大的属性,同时利用属性之间的关联度,使处理数据的范围不断缩小来减少求约简的时间。该算法在计算量减少的同时能得到更简的结果并能得到所有相对约简,实例分析也验证了该算法的有效性。
Reduction of attributes is one of the key problems in the knowledge discovery.In order to achieve the minimal relative re-duction,the paper appends the most significance of attributes to core of attributes from original set of core attributes based on the discernibility matrix.And the paper can save some time of acquiring the least reduction making use of the association degree of attributes.The algo-rithm can obtain the more reductive result requiring less computing and get all of the relative reductions.Finally,the experimental results show that this algorithm is effective.
出处
《计算机与数字工程》
2012年第4期17-18,31,共3页
Computer & Digital Engineering
关键词
属性约简
核
区分矩阵
关联度
attribute reduction
core
discernibility matrix
association degree