摘要
属性约简要求在保证分类和决策能力不变的前提下删除数据库中的冗余属性,简化知识表示,生成决策规则,从而为科学化的管理、预测和决策提供有力的支持。深入系统地研究了基于粗糙集理论的属性约简问题,分别对于相容和不相容决策表,在研究了现有约简算法的基础上,发现算法的局限性、优点和不足,并针对不足,对算法作了一些改进。
Attribute reduction can delete the redundant attributes in the database to simplify the decision table in order te help people making the right and concise decisions. In this article, I studied attribute reduction systematically and extensively. I studied the existing attribute reduction algorithm and discovery the limitations of the algorithm and the weaknesses of the algorithm. And I made some improvement.
出处
《齐齐哈尔大学学报(自然科学版)》
2013年第3期12-14,共3页
Journal of Qiqihar University(Natural Science Edition)
基金
蚌埠学院自然科学基金项目(2011ZR19)
关键词
粗糙集
属性约简
核属性
区分矩阵
rough sets
atribute reduction
core
discernibility matrix