K-均值算法聚类分析及其在人力资源管理中的应用  

K-means Clustering Analysis Algorithm and Its Application to a HRM System

作　　者：刘艳丽[1] 刘希玉[1] 孟岩[1] 祖伟 

机构地区：[1]山东师范大学管理与经济学院,山东济南250014

出　　处：《山东科学》2008年第2期33-38,共6页Shandong Science

基　　金：山东省自然科学基金重大项目(Z2004G02);泰山学者建设工程专项经费资助项目;山东省教育厅计划项目(J05G01)

摘　　要：本文提出了一种改进的K-均值聚类算法,在基本K-均值算法的基础上运用基于密度选择初始中心点并且通过学习特征权值改进聚类效果,克服了基本K-均值算法初始中心点难以确定、聚类结果不稳定的缺点;然后建立了一种基于改进的K-均值算法的人事管理系统聚类分析模型,本模型采用SQL Server 2000数据库实现并成功运用于国内一家知名软件企业的人力资源管理系统中,为该企业选聘人才和用好人才提供了有益的参考。An improved K-meaus clustering algorithm is proposed, which is based on basic K-meaus Algorithm, and which choose original clustering centers based on densities and improves clustering effects according to feature weight learning. It counterbalance the negatives of difficulty in choosing original clustering centers and of unstable clustering effects. The clustering analysis model of a human resource management system is then put forward, which is based on the improved K-means clustering algorithm. The realization of the model, which employs SQL Server 2000 is successfully applied to the human resource management of a well-known domestic software company and offers a useful reference for the enterprise to select and appoint talentes.

关 键 词：K-均值算法 密度 聚类中心 特征权值 人力资源管理 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]