Risk management, signal processing and econometrics: A new tool for forecasting the risk of disease outbreaks

• Published in: Journal of theoretical biology 2019-04, Vol.467, p.57-62
• Main Authors: Hassani, Hossein, Yeganegi, Mohammad Reza, Silva, Emmanuel Sirimal, Ghodsi, Fatemeh
• Format: Article
• Language: English
• Subjects: Disease; Forecasting; Multivariate singular spectrum analysis; Outbreaks; Quantile regression; Value at risk
• ISSN: 0022-5193; 1095-8541
• DOI: 10.1016/j.jtbi.2019.01.032

•Risk management, signal processing and econometrics based novel approach for forecasting the risk of disease emergence.•We propose risk quantification using the Value at Risk criterion.•Multivariate Singular Spectrum Analysis is combined with Quantile Regression (MSSA-QR) in a two staged model.•The MSSA-QR model exceeds in forecasting worst case scenarios for less common waterborne diseases.•Forecasting of more common diseases requires the inclusion of socio-economic and environmental indicators. This paper takes a novel approach for forecasting the risk of disease emergence by combining risk management, signal processing and econometrics to develop a new forecasting approach. We propose quantifying risk using the Value at Risk criterion and then propose a two staged model based on Multivariate Singular Spectrum Analysis and Quantile Regression (MSSA-QR model). The proposed risk measure (PLVaR) and forecasting model (MSSA-QR) is used to forecast the worst cases of waterborne disease outbreaks in 22 European and North American countries based on socio-economic and environmental indicators. The results show that the proposed method perfectly forecasts the worst case scenario for less common waterborne diseases whilst the forecasting of more common diseases requires more socio-economic and environmental indicators.