ARIMA模型在预测甘肃省梅毒月发病情况中的应用  

Using ARIMA model of syphilis incidence in every moth in Gansu Province

作　　者：李莉[1] 孟蕾[1] 王斌[1] 杨明宇[1] 苟伟斌[1] 郁华[1] 石林[1] 

机构地区：[1]甘肃省疾病预防控制中心性病艾滋病科,甘肃兰州730000

出　　处：《中国皮肤性病学杂志》2015年第2期160-163,共4页The Chinese Journal of Dermatovenereology

基　　金：甘肃卫生行业科研计划资助目(编号:GSWSKY-2014-22)

摘　　要：目的评价自回归移动平均(auto-regressive integrated moving average,ARIMA)模型在预测梅毒发病趋势中的作用,为各地制定梅毒防治策略提供科学的参考依据。方法收集甘肃省2004-2013年梅毒发病数资料,用SPSS软件拟合ARIMA模型并预测2014年梅毒月发病情况。结果拟合ARIMA(1,1,2)(0,1,1)12模型为甘肃省梅毒月发病率预测的最佳模型,残差为白噪声的短期预测,预测值与实际值平均相对误差值很低,且实际值均在预测值的95%可信区间范围,即预测结果可靠。结论 ARIMA模型能较好的预测梅毒发病的短期变化趋势,可供制定梅毒预防控制措施进行参考。Objective To forecast its tendency of syphilis incidence in Gansu with ARIMA model, which may provide a scientific proof for making the following preventive and control measure. Methods Based on the data of syphilis monthly cases during 2004 -2013 to forecast the incidence levels in 2014 with ARIMA model by SPSS. Results ARIMA （ 1,1,2） （0,1,1） 12 is the best forecasting model of syphilis monthly incidence rate model. The residual sequence is a white noise sequence of short-term prediction. The relative error in average is lower between the forecasting value and the real value. And the real value in the 95% confidence interval range of the forecasting value. The prediction result is reliable. Conclusion ARIMA model can provide the forecast in its incidence tendency of syphilis in short time, which can be used to predict the the strategies of control and prevention.

关 键 词：梅毒 ARIMA模型 预测 甘肃省 

分 类 号：R759.1[医药卫生—皮肤病学与性病学]