September 16, 2016
Evan Ray, a postdoctoral researcher in Biostatistics, recently attended a workshop on forecasting outbreaks of influenza at the Centers for Disease Control and Prevention (CDC) Roybal Campus in Atlanta, GA. Ray, who attended with Assistant Professor Nicholas Reich, developed a new method for forecasting infectious disease incidence and used this method to submit weekly forecasts for the CDC’s 2015-2016 Seasonal Influenza Forecasting Competition that started in November 2015. Their method was the top overall performer in predicting the onset and peak timing of the 2015-2016 influenza season in the US. This meeting was the concluding session for the contest. The workshop, held August 31-September 1, brought together the different teams in the competition to share their methods with each other as well as external disease modeling experts and applied epidemiologists, as well as to discuss future directions for the contest.
“It was interesting to get a better understanding of aspects of the data and disease dynamics that are relevant to making predictions. For example, how the data are collected, how reported flu incidence tends to peak around holidays like Thanksgiving and winter break, and how these predictions could be integrated into public health practice,” notes Ray.
Among the speakers was Dr. Daniel Jernigan, Director of the CDC’s Influenza Division, who gave a talk about the CDC’s forecasting and decision making processes. Other teams besides UMass Amherst included representatives from three private companies as well as Carnegie Mellon University, Columbia University, Imperial College London, Iowa State University, Northeastern University, University of Arizona, and the University of Minnesota.