基于边界区域局部模糊增强的πRKM聚类算法  

Improved πRKM clustering algorithm based on local fuzzy enhancement of boundary region

作　　者：马福民[1] 逯瑞强 张腾飞[2] 

机构地区：[1]南京财经大学信息工程学院,南京210023 [2]南京邮电大学自动化学院,南京210023

出　　处：《控制与决策》2017年第11期1949-1956,共8页Control and Decision

基　　金：国家自然科学基金项目(61403184;61105082);江苏省高校自然科学研究重大项目(17KJA120001);南京邮电大学1311人才计划基金项目(NY2013);江苏高校优势学科建设工程项目;国家电子商务信息处理国际联合研究中心项目(2013B01035)

摘　　要：如何对交叉边界区域的数据对象进行度量与处理一直是粗糙k-means(RKM)及其衍生算法的主要出发点.πRKM算法通过引入Laplace无差别原则,较好地解决了传统RKM算法对权重系数的选择比较敏感等相关问题,但没有考虑边界区域多个类簇的交叉程度以及边界区域数据对象的空间位置分布对聚类结果的影响.鉴于此,设计一种对边界区域的数据对象进行局部模糊度量的方法,并提出基于边界区域局部模糊增强的πRKM聚类改进算法,通过多组实例分析验证了所提出算法的有效性.The primary starting point of rough k-means（RKM） and its derivatives is how to measure and process the data objects in the boundary regions. The traditional RKM algorithm is more sensitive to the choice of the weight coefficients of the upper and lower approximations, and the partitioning results are easy affected by the non-competitive objects in boundary region. By introducing the Laplace＇s principle of indifference for measuring the objects in boundary regions,the aforementioned problems of the traditional RKM algorithm are sovled well by using the πRKM algorithm. However,the overlapping degree in boundary regions and spatial distributions of different boundary objects are not considered. In order to better describe data objects in boundary regions, the local fuzzy measurement is introduced, and an improvedπRKM clustering algorithm based on local fuzzy enhancement of boundary region is developed. The effectiveness of the proposed algorithm is demonstrated by experimental comparison and analysis.

关 键 词：粗糙聚类 K-MEANS 局部模糊度量 粗糙集 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]