8 EWS
We are developing a disease surveillance system to forecast the future number of cases and the early detection of disease outbreaks. We are validating the system using dengue data from Brazil.
System components
- Data retrieval of official dengue data, search query data, population, climate and environmental covariates (download, clean and format data for analysis and visualization)
- Nowcasting methods using digital, environmental and climate data. Data include temperature, precipitation, and Google trends using specific keywords related to dengue)
- Forecasting methods using environmental and climate data (and ensemble of several methods)
- Methods for monitoring dengue in one municipality, and extension to models that borrow information from neighboring regions
- Methods performance by quantifying how well the methods predict future number of cases weeks or months in advance
- Software system to be used to monitor dengue in Brazil, and adaptable to other diseases and countries