摘要
讨论模糊C均值聚类算法在决策表条件属性对决策属性的相容程度的指导下对粗集理论中的连续属性进行离散化的一种新算法。该算法充分考虑属性之间的相关性,将所有连续属性转化为矩阵同时处理,能明显提高传统动态层次聚类算法离散化过程的速度。算法测试结果表明,新算法能较好地保留有效属性,提高离散化精度。
In this article, the authors propose a new method for discretization of continuous attributes based on FCM cluster using feedback information from decision table consistency. As the interactions between attributes are considered, this algorithm transforms all the continuous attributes into matrix to process simultaneously, which can obviously speed the discretization process of traditional dynamic hierarchical clustering algorithm. Results show that this approach can not only keep the effective attribute but also increase discretization precision.
出处
《重庆邮电学院学报（自然科学版）》
2006年第5期650-652,678,共4页
Journal of Chongqing University of Posts and Telecommunications(Natural Sciences Edition)
基金
四川省教育厅应用研究基础项目资助(2005A140)
