您的位置:首页  > 论文页面

一种新的粗糙集约简算法

发表时间:2008-06-30  浏览量:2462  下载量:1060
全部作者: 肖厚国,李岚,宫悦
作者单位: 大连理工大学城市学院;大连海事大学
摘 要: 粗糙集是一种处理模糊和不确定性数据的工具,是数据挖掘中的重要框架。属性约简是粗糙集研究的核心内容之一,数据经过约简后更有价值,更能准确地获取知识。现已证明寻找最小约简是NP-hard问题。由于基于最初差别矩阵的属性约简的定义与基于正区域属性约简定义是不一致的,本文提出了一种新的区分矩阵与免疫遗传算法结合的方法,该方法能够实现相容/不相容决策表的属性约简。先简化差别矩阵,降低求不可区分关系的算法复杂度,通过简化区分函数方法求得核属性,有效地提高计算速度。改进后的算法可以求取属性的一个约简。最后通过实例证明了算法的有效性。
关 键 词: 应用数学;粗糙集;免疫遗传算法;区分矩阵;属性约简
Title: A new reduction algorithm based on rough set theory
Author: XIAO Houguo, LI Lan, GONG Yue
Organization: City Institute, Dalian University of Technology;Dalian Maritime University
Abstract: Rough set theory is a tool to deal with vague and uncertain data, and it becomes an important frame in data mining. Attribute reduction is one of the key topics of rough set theory. Reducted data are more valuable and capable of obtaining accurate knowledge. Search for minimum reduction has been proved to be a NP-hard problem. As the definition of attribution reduction based on primary discernibility matrix is not the same as the definition of attribution reduction based on positive region, a new rough set reduction algorithm based on immune genetic algorithm is proposed in this paper. With a new type of discernibility matrix adopted, the algorithm exhibits excellent capabilities in attribute reduction of inconsistent decision table. Simple discernibility matrix is improved so that the computing complexity of indiscernibility relation is cut down. By simplifying binary discernibility matrix and correspondence, the core algorithm is computed and the computing speed is improved efficiently. Finally, the validity and feasibility of the algorithm is demonstrated by classical databases.
Key words: applied mathematics; rough set; immune genetic algorithm; discernibility matirx; attribute reduction
发表期数: 2008年10月第12期
引用格式: 肖厚国,李岚,宫悦. 一种新的粗糙集约简算法[J]. 中国科技论文在线精品论文,2008,1(12):1343-1347.
 
3 评论数 0
暂无评论
友情链接