Comparative Study of Four Time Series Methods in Forecasting Typhoid Fever Incidence in China

Xingyu Zhang,Min Yang,Tao Zhang,Xiaosong Li,Alistair A Young,Yuanyuan Liu,Alessandro Vespignani

doi:10.1371/journal.pone.0063116

Abstract

Accurate incidence forecasting of infectious disease is critical for early prevention and for better government strategic planning. In this paper, we present a comprehensive study of different forecasting methods based on the monthly incidence of typhoid fever. The seasonal autoregressive integrated moving average (SARIMA) model and three different models inspired by neural networks, namely, back propagation neural networks (BPNN), radial basis function neural networks (RBFNN), and Elman recurrent neural networks (ERNN) were compared. The differences as well as the advantages and disadvantages, among the SARIMA model and the neural networks were summarized and discussed. The data obtained for 2005 to 2009 and for 2010 from the Chinese Center for Disease Control and Prevention were used as modeling and forecasting samples, respectively. The performances were evaluated based on three metrics: mean absolute error (MAE), mean absolute percentage error (MAPE), and mean square error (MSE). The results showed that RBFNN obtained the smallest MAE, MAPE and MSE in both the modeling and forecasting processes. The performances of the four models ranked in descending order were: RBFNN, ERNN, BPNN and the SARIMA model.

Highlights

Typhoid fever is a disease caused by the bacterium, Salmonella enteric subspecies enteric serovar Typhi, and is common in developing and underdeveloped countries [1]
Comparative studies of different forecasting techniques can facilitate the selection of the best time series model for forecasting future epidemic behavior in specific types of diseases [24]. We address this problem by comparing the forecasting performance of the seasonal autoregressive integrated moving average (SARIMA) model and three typical artificial neural networks, namely, back propagation neural networks (BPNN), radial basis function neural networks (RBFNN) and Elman recurrent neural networks (ERNN) in shortterm forecasting for typhoid fever, using typhoid fever incidence data for Guangxi province, China, for illustration
autoregressive integrated moving average (ARIMA) and SARIMA models have been widely used for epidemic time series forecasting including the hemorrhagic fever with renal syndrome [8,58], dengue fever [9,59], and tuberculosis [10]

Summary

Introduction

Typhoid fever is a disease caused by the bacterium, Salmonella enteric subspecies enteric serovar Typhi, and is common in developing and underdeveloped countries [1]. According to the WHO, an estimated 22 million cases of typhoid fever occur annually, with at least 200,000 deaths. In some underdeveloped areas of China, typhoid fever is still a serious infectious disease that severely affects lives of the patients. The need arises for a modeling approach that can provide decision makers early estimates of future typhoid fever incidence based on the historical time series data. The goal is to monitor and predict the trends in typhoid fever incidence to facilitate early public health responses to minimize morbidity, mortality and the adverse clinical outcomes of the patients

Methods

Results

Conclusion