粗糙集理论中的离散化问题

Discretization in Rough Set Theory

一引言数据分析及数据挖掘,是一个重要的正在迅速发展的研究课题。波兰科学家Z.Pawlak于1982年提出的粗糙集(Rough Set)理论正是解决这一问题的新理论,它可用于处理决策信息表中的不确定知识,并用规则的形式表达,是一种有效的知识获取工具。在Roughset中,数据约简是非常重要的,包括属性约简和值约简。运用粗糙集理论处理决策表时。

In this paper,the discretization problem in Rough Set is discussed. We do classification research on the discretization algorithm from different viewpoints. Several discretization algorithms are introduced and discussed. We improve the greedy algorithm in different ways. We present a strategy for discretization based on the importance of condition attributes. In the end, we present the goal of discretization based on experiments.