粗糙集与粒子群算法结合的属性离散化方法  

Method of discretization of continuous attributes based on rough set and Particle Swarm Optimization

作　　者：沈中林[1] 范敬德[1] 樊玮[1] 

机构地区：[1]中国民航大学计算机学院,天津300300

出　　处：《计算机工程与应用》2009年第12期129-130,159,共3页Computer Engineering and Applications

基　　金：国家自然科学基金委员会与中国民用航空总局联合资助项目(No.60672173)

摘　　要：提出了一种结合粗糙集和粒子群的连续属性离散化算法,采用了MPSO算法的思想,提高了粒子群摆脱局部极值的能力,得到了较好的离散化效果。对不同的数据集进行了多次测试,结果表明该算法在对数据离散化时有较好的性能。This paper presents an algorithm for discretization based on rough set and particle swarm optimization.h mainly inducts a modified PSO （MPSO） and improves the ability to break away from the local optimum.Simultaneously,the algorithm obtains a satisfactory discretization result.Authors do many tests for different data sets and the experimental results prove that the algorithm has better validity for discretization of continuous attributes.

关 键 词：粒子群优化 粗糙集 离散化 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]