基于ARIMA模型的广西新型冠状病毒肺炎疫情发展趋势预测  

Development trend prediction of COVID-19 epidemic in Guangxi based on ARIMA model

作　　者：刘忠典 黎燕宁[1] Liu Zhongdian;Li Yanning(Department of Biostatistics,School of Public Health,Guangxi Medical University,Nanning 530021,China)

机构地区：[1]广西医科大学公共卫生学院统计学教研室,南宁530021

出　　处：《广西医科大学学报》2021年第12期2367-2371,共5页Journal of Guangxi Medical University

摘　　要：目的:分析广西新型冠状病毒肺炎(COVID-19)确诊病例的时空特征,预测COVID-19确诊病例的动态变化趋势,为今后类似的新发传染病疫情暴发时,对其流行病学分布、发展趋势及预警防控提供科学依据和借鉴意义。方法:从时间和空间方向分析研究2020年1—2月广西全区COVID-19确诊病例的流行特征,构建自回归求和移动平均(ARIMA)模型。结果:全区中疫情最为严重的地级市有南宁市、桂林市和北海市,其他市疫情较轻;确诊人数初期增长较快,2月中旬趋于稳定,进入平稳缓慢期;ARIMA(0,1,0)模型显示,预测值与真实值拟合度较高。结论:广西COVID-19流行在一个月内得到有效控制;ARIMA(0,1,0)具有较好的拟合效果,可用于COVID-19流行趋势的短期预测。Objective:To analyze the spatial and temporal characteristics of the confirmed cases of coronavirus disease 2019(COVID-19)in Guangxi,and to predict the dynamic change trend of the confirmed cases of COVID-19,so as to provide scientific basis and reference for the epidemiological distribution,development trend,early warning and prevention and control of similar outbreaks of emerging infectious diseases in the future.Methods:The epidemiological characteristics of confirmed COVID-19 cases in Guangxi from January to February in 2020 were analyzed from the perspective of time and space.Building the autoregressive integrated moving average model(ARIMA).Results:Nanning,Guilin and Beihai were the most severely affected prefecture-level cities in the whole region,while other cities were less affected.The number of diagnosed patients increased rapidly in the early stage,and tended to be stable in the middle of February.The ARIMA(0,1,0)model showed that the predicted value fitted well with the real value.Conclusion:The prevalence of COVID-19 in Guangxi is effectively controlled within one month.ARIMA(0,1,0)has a good fitting effect and can be used for short-term prediction of COVID-19 trend.

关 键 词：新冠肺炎 自回归求和移动平均模型 疫情 预测 

分 类 号：R183.3[医药卫生—流行病学]