一种用于短期肺气肿发病率预测的分时段时间序列传递函数模型  

Pulmonary Emphysema Incidence Rate Forecasting Model Based on Transfer Function Models for Period-Decoupled Time Series

作　　者：马亮亮[1] 田富鹏[1] 

机构地区：[1]西北民族大学计算机科学与信息工程学院,甘肃兰州730030

出　　处：《药物生物技术》2010年第5期408-411,共4页Pharmaceutical Biotechnology

基　　金：国家自然科学基金资助项目(No.60673192)

摘　　要：肺气肿发病率是政府和相关医学工作者预防与监测肺气肿的重要依据之一。因此,选择合理的肺气肿发病率预测模型显得尤为重要。文章提出了一种基于时间序列法的分时段传递函数模型来预测短期肺气肿发病率,该模型考虑了平均气温因素对肺气肿发病率的影响,同时利用累积式自回归滑动平均模型(ARIMA)对肺气肿发病率序列和平均气温序列的非平稳性进行处理,并且对2009年12个时段分别建立了预测模型。最后采用青海海西州地区的肺气肿发病率历史数据进行算例研究。结果表明,利用本文模型进行肺气肿发病率预测能够提高预测的准确性。对肺气肿发病率历史数据进行时间序列分析是用于肺气肿监测的一个重要的内容。该研究所建立的传递函数预测模型适用于青海海西州地区肺气肿发病率预测与监测。Pulmonary emphysema incidence rate is an important basis for government and medic to prevent and to survey pulmonary emphysema,and so selecting the equitable forecasting model of pulmonary emphysema incidence rate seems very important.A pulmonary emphysema incidence rate forecasting model based on Transfer Function Models for period-decoupled time series was presented in this paper.The effect of average temperature on the pulmonary emphysema incidence rate can be fully taken into account.The ARIMA model is employed to deal with the non-stationary series of pulmonary emphysema incidence rate and average temperature,and forecasting models for every month＇s period of 2009 are developed independently.The numerical example is based on Qinghai Haixizhou region pulmonary emphysema incidence rate data.The results showed that the proposed model could improve the accuracy of forecasting.Time series methods applied to historical reporting data of pulmonary emphysema incidence rate are an important tool for pulmonary emphysema surveillance.The established transfer function forecasting model is suitable to forecast the incidence rate of pulmonary emphysema in Qinghai Haixizhou region.

关 键 词：ARIMA模型 传递函数模型 预测 发病率 时间序列 肺气肿 

分 类 号：R181.8[医药卫生—流行病学]