一种改进BP神经网络的K-means算法  

A K-Means Algorithm for Improving BP Neural Network

作　　者：王伟 储泽楠[1,3] 张修太[2] 韩毅[1,4,5] WANG Wei;CHU Zenan;ZHANG Xiutai;HAN Yi(School of Computer Science and Information Engineering,Anyang Institute of Technology,Anyang He’nan 455000,China;School of Electronic Information and Electrical Engineering,Anyang Institute of Technology,Anyang He’nan 455000,China;Anyang Institute of Technology Anyang Information System Application Engineering Technology Research Center,Anyang He’nan 455000,China;Anyang Institute of Technology Henan Province High Precision Spindle Engineering Laboratory,Anyang He’nan 455000,China;National CNC System Engineering Technology Research Center,Huazhong University of Science and Technology,Wuhan 430000,China)

机构地区：[1]安阳工学院计算机科学与信息工程学院,河南安阳455000 [2]安阳工学院电子信息与电气工程学院,河南安阳455000 [3]安阳工学院安阳市信息系统应用工程技术研究中心,河南安阳455000 [4]安阳工学院河南省高精密主轴工程实验室,河南安阳455000 [5]华中科技大学国家数控系统工程技术研究中心,武汉430000

出　　处：《电子器件》2020年第2期380-385,共6页Chinese Journal of Electron Devices

基　　金：河南省科技计划项目(182102210197);河南省教育厅一般项目(2020-ZDJH-002);安阳市科技发展计划项目(201804)。

摘　　要：针对聚类算法K-means在聚类的过程中,出现数据的属性缺失造成聚类的精确度下降等问题,提出了一种改进BP神经网络的BPK-means算法。该算法首先根据BP神经网络对缺失的属性值进行预测,补全缺失的数据,极大的提高了数据的完整性和可靠性;然后对异常的数据进行去噪处理;最后通过K-means算法对修复后的数据进行聚类。通过理论验证和实验结果都表明所提出的方法比原始的算法精确度有很大提高。In the process of clustering algorithm K-means,the loss of clustering accuracy caused by the loss of data attributes is proposed. A BPK-means algorithm based on BP neural network is proposed. The algorithm firstly predicts missing attribute values based on BP neural network,and completes the missing data,which greatly improves the integrity and reliability of the data;then denoises the abnormal data;finally passes the K-means algorithm. Cluster the repaired data. Both theoretical verification and experimental results show that the proposed method is much more accurate than the original one.

关 键 词：BP神经网络 属性缺失 K-MEANS 降噪处理 数据完整性 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]