IKMQT-一种改进的K-Mean算法及其在医院预约分诊排队中的应用  

IKMQT-An Improved K-Means Algorithm and the Application of Reservation-triage-queuing in Hospital

作　　者：姜永增[1] 姚仲敏[1] 张裕[1] 董晶[1] 

机构地区：[1]齐齐哈尔大学计算机与控制工程学院

出　　处：《科技通报》2012年第8期126-128,共3页Bulletin of Science and Technology

基　　金：黑龙江省教育厅科技项目基金资助(NO.12511614)

摘　　要：K-Means算法中初值K的选择直接影响到算法的效率,传统的估值方法不得不多次调整K的值,存在很大缺陷。本文提出了一种基于排队论思想指导的改进的K-Means算法IKMQT,使用排队论指导计算初始值K,同时考虑结点的资源消耗和位置信息。经过分析,IKMQT比K-Means降低了计算复杂度,通过在医院分诊排队当中的实际应用证明IKMQT算法提高了医院诊疗的效率。How to select the initial K impacts on the efficiency of K-Means,It was had to repeatedly adjust the value of K in the traditional method,which was a big defect.IKMQT,an improved K-Means algorithms by queuing theory is put forward in this paper.To calculate the initial K with direction of queuing theory,to take into account the consumption of resources and location information of the node at the same time.IKMQT reduced the computation complexity than K-Means,and IKMQT improved the efficiency of diagnosis and treatment through the practical application of triage and queuing in hospital.

关 键 词：K-MEANS 排队论 IKMQT 分诊排队 

分 类 号：TP399[自动化与计算机技术—计算机应用技术]