Prediction of Incidence Trend of Influenza-Like Illness in Wuhan Based on ARIMA Model

Abstract

Objective. The autoregressive integrated moving average (ARIMA) model has been widely used to predict the trend of infectious diseases. This paper is aimed at analyzing the application of the ARIMA model in the prediction of the incidence trend of influenza-like illness (ILI) in Wuhan and providing a scientific basis for the prediction and prevention of influenza. Methods. The weekly ILI data of two influenza surveillance sentinel hospitals in Wuhan City published on the website of the National Influenza Center of China were collected, and the ARIMA model was used to model the data from 2014 to 2020, to predict and verify the ILI data in 2021. Results. The optimal model for the incidence trend of ILI in Wuhan was ARIMA 1 , 1 , 1 , the residuals were in line with the white noise sequence ( 0.018 < Ljung ‐ Box Q < 30.695 , P > 0.05 ), and the relative error between the predicted value and the actual value was small, which all proved the model was practical. Conclusion. ARIMA 1 , 1 , 1 can effectively simulate the short-term incidence trend of ILI in Wuhan.

Predicting the epidemiological trend of acute hemorrhagic conjunctivitis in China using Bayesian structural time–series model

“…The ARIMA model predicts future changes from past and present values by describing the autocorrelation of random variables dependent on time 31 . According to whether the original sequence of the model is stationary and the different parts contained in the regression, it can be divided into autoregressive model AR (p), moving average model MA (q), autoregressive moving average process ARMA (p, q) , and ARIMA (p, d, q) 14,31 . When the original series contains seasonal trends, the seasonal ARIMA model is used, and its basic structure is ARIMA (p, d, q) (P, D, Q).…”

“…When the original series contains seasonal trends, the seasonal ARIMA model is used, and its basic structure is ARIMA (p, d, q) (P, D, Q). d and D represent ordinary difference and seasonal differencing orders, respectively; p and q represent the autoregressive order and moving average order in the model, respectively; P and Q are the order of the seasonal model autoregression and moving average, respectively, and s is the seasonal cycle 14 . There are typically four steps 10 in building the model (including identification, estimation, diagnosis, and prediction), allowing the optimal model parameters to be automatically selected in the ARIMA builder.…”

“…Since then, the ARIMA model has been widely used in various fields such as economy, industry, and transportation, and it has played an important role in the prediction and early warning of infectious diseases [11][12][13] . However, the sequence value of the ARIMA model is affected by various factors such as time, resulting in irregular changes in the sequence, so that the original fitting model is no longer applicable or the prediction performance is reduced 14 .…”

Model-based analysis of the incidence trends and transmission dynamics of COVID-19 associated with the Omicron variant in representative cities in China

“…The ARIMA model can be applied to capture fluctuations in historical data, and the model itself is constructed with the help of endogenous variables, so it is well suited to address the interference of external factors. Furthermore, the model has the advantages of a simple structure, high applicability, and strong data interpretation ability and is widely used in the short-term prediction of infectious diseases [ 13 , 14 ]. The LSTM model is a type of deep learning network that is able to “recall” patterns in past or future data without artificially adding temporal features and can explore the nonlinear correlation characteristics between time series data to a large extent [ 15 , 16 ].…”

Forecasting Influenza-Like Illness (ILI) during the COVID-19 Pandemic

“…Participants in the FluSight challenge have implemented and assessed a wide variety of methods for forecasting influenza-like illness (ILI), including compartmental models, statistical (e.g., time series) models, and machine learning approaches. Infectious disease modelers have demonstrated the utility of time series methods for forecasting ILI, and the performance of specific approaches such as autoregressive integrated moving average (ARIMA) and autoregressive neural network (AR-NN) has been studied previously (Kandula and Shaman 2019a; Meng, Huang, and Kong 2022; Tsan et al 2022).…”