基于K-means聚类算法的人事管理异常数据识别和自动处理系统  

Personnel management abnormal data recognition and automatic processing system based on K-means clustering algorithm

作　　者：韩晓萃 胡业维 吴庆艳[4] 胡敏 曾思颖 HAN Xiaocui;HU Yewei;WU Qingyan;HU Min;ZENG Siying(Guangdong University of Finance&Economics,Guangzhou 510060,China;The Affairs Center of Health Commission of Guangdong Province,Guangzhou 510060,China;South China University of Technology,Guangzhou 510060,China;Guangdong Pharmaceutical University,Guangzhou 510060,China)

机构地区：[1]广东财经大学,广东广州510060 [2]广东省卫生健康委员会事务中心,广东广州510060 [3]华南理工大学,广东广州510060 [4]广东药科大学,广东广州510060

出　　处：《电子设计工程》2024年第24期27-31,共5页Electronic Design Engineering

基　　金：广东省科技计划项目(2024B1212070015);广东省医学科学技术研究基金(B2023284,C2024052,C2022248)。

摘　　要：针对人事管理异常数据影响人事管理水平的问题,设计基于K-means聚类算法的人事管理异常数据识别和自动处理系统。利用全局优化K-means聚类算法,对人事管理数据进行聚类处理。该算法搜寻高密度的人事管理数据作为初始聚类中心,将人事管理数据聚类为多个簇。利用K-means聚类算法构建人事管理数据的自回归模型,确定人事管理数据参量的转移概率序列。转移概率序列非聚类簇中的数据时,对应数据即为人事管理异常数据识别结果。采用指数加权移动平均数方法自动修正处理所识别的人事管理异常数据。系统测试结果表明,所设计系统能够有效识别人事管理考勤数据、薪资数据中的异常数据,能够自动修正异常数据,使人事管理数据恢复正常。In response to the problem of abnormal data in personnel management affecting the level of personnel management,a personnel management abnormal data recognition and automatic processing system based on K-means clustering algorithm is designed.Utilize global optimization K-means clustering algorithm to cluster personnel management data.This algorithm searches for high-density personnel management data and uses it as the initial clustering center to cluster the personnel management data into multiple clusters.Using the K-means clustering algorithm,construct an autoregressive model for personnel management data and determine the transition probability sequence of personnel management data parameters.When transferring probability sequences to non clustered data,the corresponding data is the identification result of abnormal personnel management data.Using the index weighted moving average method,automatically correct and process identified abnormal personnel management data.The system test results show that the designed system effectively identifies abnormal data in personnel management attendance data and salary data,can automatically correct abnormal data,and restore personnel management data to normal.

关 键 词：K-MEANS聚类算法 人事管理 异常数据识别 自动处理系统 聚类中心 转移概率 

分 类 号：TN014[电子电信—物理电子学]