선택한 단어 수는 입니다.
최소 단어 이상 선택하여야 합니다.
최소 단어 이상 선택하여야 합니다.
선택한 단어 수는 30입니다.
최대 10 단어까지만 선택 가능합니다.
최대 10 단어까지만 선택 가능합니다.
제안 방법
- In performing this procedure, we first remove a seasonal effect and stochastic trend by differencing the time series, and then apply model selection criteria such as Akaike’s information criterion (AIC) to find an optimal SARIMA model.
- In this study, we employed a SARIMA model, named SIMA(52), to predict the influenza epidemics and compared its performance with the ARGO method based on Google searches, and demonstrated that our method outperforms the ARGO. Reflecting one year’s seasonal effect of the historical ILI activity plays a key role in our analysis: the twice differencing also helps stabilize the time-varying variances.
- Although the SIMA(52) demonstrates an outstanding performance somewhat superior to that of the ARGO, it still has the same defect as the ARGO has because the CDC reports ILI activity level one∼two weeks after the target date, and henceforth, make an information gap, which seems inevitable as far as only a weekly data is available for prediction. This study reminds the practitioners of the importance of classical methods in advance of adopting a trendy one. Both methods have their own merit, and therefore, a great care is necessary when implementing them for actual usage.
이론/모형
- To overcome these defects, we instead propose to employ a classical time series model, i.e. the seasonal autoregressive integrated moving average (SARIMA) model which only uses the CDC’s ILI reports.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.