摘要： 在信息等价矩阵的基础上利用粗集理论扩展了矩阵算法,设计了相对核和相对约简以及规则获取算法,提出了增量式规则挖掘的信息压缩矩阵算法。实现了在原有规则集的基础上进行规则和规则参数的增量更新,避免了重复遍历信息向量,降低了算法的时、空复杂度。用实例证明该算法是可靠有效的,为信息系统的规则约简、获取和信息压缩提供了新的思路。

关键词: 数据挖掘, 粗集理论, 信息压缩矩阵, 增量式学习

Abstract: The rules acquisition algorithm based on information equivalence matrix is extended using rough set. Relative core and relative reduction are defined, incremental rules mining based on information compression matrix algorithm is proposed. The rules and rules’ parameters updating on the basis of existing rules are realized. It avoids traversing all information vectors repeatedly and decreases the time and space complexity. A new thought for rules acquisition, reduction and information compression is provided to further study information system.

中图分类号: 

• TP311.13