文档介绍：基于粗糙集理论的规则提取算法
摘要
粗糙集理论的主要思想是在保持分类能力不变的前提下,通过属性约简和值约简,提取决策规则。本文主要是提出了利用隶属度函数进行值约简的同时提取决策规则的算法。利用该算法可在不求得核值表的情况下,直接找到各规则的最小条件属性集,获得决策表的所有决策规则。
关键词:粗糙集;隶属度函数;规则提取
An Algorithm for Rule Extraction Based on Rough Set Theory
Abstract
The main idea of rough set theory is to extract decision rules by attribute reduction and value reduction in the premises of keeping the ability of classification. In this paper, an algorithm on value reduction, and for extracting decision rule based on the membership function is proposed. All the decision rules on decision table and the minimal rule set of reduced condition attribute set without core-valued table would be attained by this algorithm.
Keywords:rough set;membership function;rule extraction
引言
粗糙集理论是一种刻画模糊的、不完整性和不确定性的数学工具。它的主要思想是是在保持分类能力不变的前提下,通过属性约简和值约简,导出决策规则。但在目前研究的粗糙约简算法中,大多都比较关注属性的约简,把求得最佳属性约简作为设计目标。但在很多的实际应用中,有时我们并不是特别关注属性约简,而只关心求得用户所需的决策规则。
以文献[1] 中全球变暖的决策表为例,如表1所示,其中Solar energy,Volcanic activity,Residual CO2 为条件属性,Temperature为决策属性,Days count是每一个对象在一年中出现的频次。
Fact
Solar energy
Volcanic activity
Residual CO2
Temperature
Days count
1
Medium
High
Low
High
20
2
High
High
High
High
30
3
Medium
Low
High
High
90
4
Low
Low
Low
Low
120
5
High
High
Medium
High
70
6
Medium
Low
High
Low
34
利用可辨矩阵进行约简[2]可知,属性约简为{Solar energy,Volcanic activity},{Solar energy,Residual CO2},{Volcanic activity,