摘要
目的探讨异方差性时间序列模型在传染病疫情数据分析中的应用。方法分别采用ARIMA和AR-GARCH模型对某市淋病发病率月报数据进行建模和拟合。结果本资料构成的时间序列经检验具有明显异方差性,经模型比较和筛选,AR(1)-GARCH(0,1)模型能够较好的拟合本研究中传染病疫情时序数据。结论AR-GARCH模型适用于传染病疫情数据构成的异方差性时序数据分析。
Objective To explore the application of heteroscedastic time series model to the analysis of data of infectious diseases. Methods ARIMA and AR-GARCH models were used to fit the incidence of gonorrhea. Results The time series in this study, which was heteroscedastic significantly, finally was well fitted by AR （ 1 ）-GARCH （0,1） model through model selecting. Conclusions AR-GARCH model is suitable for analyzing heteroscedastic time-series data of infectious diseases.
出处
《中华疾病控制杂志》
CAS
2009年第3期355-357,共3页
Chinese Journal of Disease Control & Prevention
关键词
AR-GARCH模型
时序数据
异方差
传染病
AR-GARCH model
Time-series data
Heteroseedasticity
Communicable diseases
