银行ATM设备业务总量的时序特征分析及预测  

Research and Prediction on Time-Sequence Characteristics of the Total Banking Automatic Teller Machine Business

作　　者：韦金香 张建同[1] 

机构地区：[1]同济大学经济与管理学院,上海200092

出　　处：《上海管理科学》2017年第6期25-28,共4页Shanghai Management Science

摘　　要：本文旨在分析银行ATM设备业务总量的时序特征,并据此对其进行预测。首先,本文以十分钟为间隔,绘制了银行ATM设备业务总量的30天趋势图,发现其以日为单位,呈现出显著的周期性、扰动性和多峰性,因此本文建立了以日为周期的ATM设备业务总量时序分布模型。在求解模型的过程中,本文利用模拟退火算法将每日银行系统ATM设备业务总量按其特征分为八段,消除了业务总量时间序列的多峰性。在此基础上,建立了Holt-Winters模型对业务总量进行预测,最后用第一时段进行验证,得到95%置信区间内的预测值。本文的研究结果为银行数据监控中心判断设备运行状态提供了依据。bThis paper aims to analyze the time-sequence characteristics of the total banking ATM equipment business and predict the number of the business.Firstly,we map the banking ATM equipment 30 days of total trend diagram by taking the data of ten minutes interval.It is found that it has a significant periodicity,perturbation and multi-peak.Then we set up a time-sequence distribution model of total ATM equipment business base on the data of day interval and solve the mode by simulated annealing algorithm（SA）.We divided the total amount of ATM equipment of the daily banking system into eight segments according to its characteristics,and eliminating the multi-peak of the total time series of the total business.Finally we forecast the total volume of the business through the Holt-Winters model and use the first period data for verification.And we get the 95% confidence interval of the first period business.The results of this paper provide a basis for judging the operation status of the equipment in the bank data monitoring center.

关 键 词：时序分布模型 模拟退火 Holt-Winters模型 运行状态 

分 类 号：F832[经济管理—金融学]