海量数据分类中的模糊区域判定算法研究  

Research on the Algorithm of Fuzzy Region Decision in Mass Data Classification

作　　者：欧立奇 何媛[1] 李云飞[1] 赵郁园 刘瀚[1] OU Li-qi;HE Yuan;LI Yun-fei;ZHAO Yu-yuan;LIU Han(Department of Trade and Technology/Xijing University,Xi’an 710123,China)

机构地区：[1]西京学院商贸技术系,陕西西安710123

出　　处：《山东农业大学学报（自然科学版）》2018年第2期335-338,共4页Journal of Shandong Agricultural University：Natural Science Edition

摘　　要：针对传统的K-means聚类算法在海量数据分类中,对数据模糊分解区域判定不敏感,容易陷入局部优化解,无法实现海量数据的模糊聚类的问题,本文设计一种用于海量数据分类的模糊区域判定算法,通过计算大数据模糊区域状态特点的参量,进行大数据最优聚类中点的选取,在聚类中心区域创建任意下单个数据的互通,对每个微小数据特征实现互通;通过计算隶属度以及确定聚类数的方式,解决被隐藏在海量数据中的模糊区域真实边界,实现对海量数据模糊区域的聚类求解。结果说明,所提算法能得到精确的模糊聚类效果,提高海量数据的模糊聚类效率。Aiming at the massive data classification in the traditional K-means clustering algorithm,the fuzzy data decomposition to determine the region is not sensitive,easy to fall into local optimal solution,unable to realize fuzzy massive data clustering problem,algorithm design for data classification by fuzzy region,parameter calculation of fuzzy state regional characteristics of big data,select the optimal clustering point of big data,in the clustering center area to create any single data exchange,data of each tiny feature to achieve interoperability by calculating the degree of membership and the way to determine the number of clusters,solve the hidden fuzzy region real boundary in the massive data and realize clustering is fuzzy region of massive data.The experimental results show that the proposed algorithm can obtain accurate fuzzy clustering results and improve the fuzzy clustering efficiency of mass data.

关 键 词：数据分类 模糊区域判定 算法 

分 类 号：R857.3[医药卫生—航空、航天与航海医学]