XPRIZE winners in Collective Pandemic Intelligence
© 2020 EPFL Intelligent Global Health (iGH)
The Intelligent Global Health (iGH) group hosted at the Machine Learning and Optimization Laboratory in EPFL are winners of the XPRIZE in Collective Pandemic Intelligence for their "What if…?" Global COVID policy simulator, which seeks to predict the optimal mix of COVID-19 mitigation policies tailored to a specific time and place.
As part of his MSc thesis, Thierry Bossy developed a deep learning model using recurrent neural networks to assess the country-specific epidemiological utility of COVID-19 mitigation policies. The model is fed various global datasets on mobility, weather and socio-economic features. Research interns, Lucas Massemin and Kimia Hemmatirad, further harness epidemiological trends from social media while Dr Tatjana Chadarova(MLO Postdoc) and MSc student, Pablo Canas are working on deep learning interpretability.
The model is now being incorporated into an open-source open access web-based platform by BSc student Andrea Pinto to act as a communication tool between the public and policymakers to make policies transparent, trusted and tolerated.
The project is co-supervised by Dr Mary-Anne Hartley (head of iGH), Prof Martin Jaggi and Prakhar Gupta (PhD student).