欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
Dubois粗糙模糊集中，上、下近似集的隶属度分别由等价类中元素隶属度的上、下确界来确定，由于没有充分反映出等价类中隶属度介于上、下确界之间的那些元素的作用，在信息处理中不免造成这些元素信息的丢失。为此提出一种新的粗糙模糊集近似算子的表示方法，该方法能够涵盖等价类中所有元素的隶属度。阐述了该方法在信息处理中的合理性，给出了相应的代数性质；在此基础上重新定义了粗糙模糊隶属函数；给出新算子下近似分类的精度、分类质量、属性的依赖度及基于依赖度的属性约简算法；最后用实例说明了算法的有效性。
英文摘要：
For Dubois rough fuzzy sets, the membership of the lower or upper approximations is defined as the elements memberships＇ infimum or supsmum in equivalent class. As a result of not considering the elements whose memberships are between the minimum and maximum, some useful information of these elements may be lost in the information processing. The paper presents a new operator of rough fuzzy sets that every element＇s membership in equivalent class is taken into account. Based on the new operator, algebra properties are put forward and rough fuzzy membership is defined. More-over, the paper presents accurate degree, classified quality, dependence degree and attribute reduction algorithm. At last, an example proves that the algorithm is efficient.
同期刊论文项目
同项目期刊论文
期刊信息
位置：