论文部分内容阅读
知识约简是知识发现中的一项重要任务,它的研究使得在数据中寻找蕴含规则更加简易,但现实中的信息往往都不是一成不变的,随着时间的推移需要不断地对原有信息进行适当更新。本文主要从粒辨识属性矩阵出发,讨论当所研究的形式背景处于更新状态时,如何由原有的粒协调集更新得到新的粒协调集,并讨论了粒辨识属性矩阵的相关性质。
Knowledge reduction is an important task in knowledge discovery. Its research makes it easier to look for implicative rules in data. However, the information in reality is not always static, and as time goes on, the original information needs to be constantly Appropriate update. In this paper, starting from the granule identification matrix, we discuss how to update a new granule coordination set from the original granule coordination set when the formal context under study is updated, and discuss the related properties of the granule identification attribute matrix.