论文部分内容阅读
目前许多挖掘算法都试图使异常信息的影响最小化,或者排除它们,经典粗糙集理论基于正域的属性约简方法也不例外,它直接排除了边界域中样本所包含的信息.如何改变边界域结构,将边界域样本尽可能拓展到正域结构中,从而有效获取更有价值信息的研究很有必要.在经典粗糙集理论的基础上,采用统计学中基于某种偏好策略,提出了边界域重构的基本方法和知识获取方法,进一步讨论了与变精度粗糙集模型之间的联系,并重新定义了变精度粗糙模型中关于β下近似的定义.结果表明基于边界域重构的方法和修正后变精度粗糙集模型正域结构得到相应扩大,获取异常信息能力进一步加强.
At present, many mining algorithms try to minimize or eliminate the influence of abnormal information. The classical rough set theory based on the positive domain attribute reduction method is no exception, it directly excludes the information contained in the sample of the boundary domain .How to change the boundary Domain structure, it is necessary to extend the sample of the boundary domain to the positive domain structure as much as possible so as to effectively obtain more valuable information.On the basis of classical rough set theory, based on some preference strategy in statistics, The basic method and knowledge acquisition method of boundary domain reconstruction are discussed in detail, and the relation between them and the variable precision rough set model is further discussed and the definition of lower approximation in variable precision rough model is redefined.The results show that based on the boundary domain reconstruction The positive domain structure of the method and the modified precision-rough set model are expanded correspondingly, and the capability of acquiring abnormal information is further strengthened.