COMBATCOVID19

The COVID 19 pandemic imposed enormous costs on society and the economy. A crucial issue concerns which types of interventions should be implemented to effectively curb its spread, prevent future waves, and minimize negative impacts. Designing effective interventions requires a deep understanding of the transmission network across the different regions of the country. We therefore proposed a multidisciplinary, integrated, data driven modeling approach to support policymakers in the assessment and control of the pandemic in Italy. Novel machine learning and manifold learning methods were developed for large scale epidemic time series forecasting, along with data driven numerical techniques that solve ordinary and partial differential equation models more accurately and efficiently than traditional approaches.