偏好关系下的最优约简求解算法研究  

Research of optimal reduct under preference

作　　者：张小峰[1] 张福增[1] 李明[2] 王娜[2] 

机构地区：[1]烟台师范学院计算机科学与技术学院,山东烟台264025 [2]兰州理工大学计算机与通信学院,甘肃兰州730050

出　　处：《计算机工程与设计》2005年第8期2103-2106,共4页Computer Engineering and Design

摘　　要：随着数据资源为越来越多的用户所共享,不同用户对相同数据信息有不同的用户偏好与需求,并由此而产生了一种数据挖掘技术,称之为面向需求的知识发现。针对粗糙集中的约简求取问题,提出了字典序偏好关系,并精确地描述了此偏好关系下的最优约简,设计了实现算法的数据结构及相应的访问方式FLVA,构造出了相应的求解算法,并从理论上证明了该算法的正确性。With more and more users sharing with the same data resource, different users have different preference and requirement to the same data information. Based on this, a new data mining technique emerges, which is called requirement oriented knowledge discovery. According to the problem of retrieving reducts of rough sets, a special preference relation is presented, which is dictionary relation. The optimal reduct under such preference relation was described, and the data structure was constructed to solute the problem and the corresponding algorithm on the designed data structure. At last and the algorithm to retrieve the optimal reduct under the preference relation was constructed, and the correctness of the algorithm theoretically is proven.

关 键 词：偏好 字典序 约简 FLVA ROUGHSET 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]