A compartmental model for the analysis and prediction of COVID 19 spread

Mathematical models can help to understand natural phenomenas and to predict their dynamics. In this thesis we try to apply them to the COVID-19 disease. We briefly present the fundamental ideas that underlie the activity of modelling and more specifically the main concepts of modeling infectious diseases. We then propose a forced SEIRD differential model and we analyze and forecast the COVID-19 spread in Emilia Romagna and Lombardia, using the Data provided by the Italian Civil Protection Department. In this study we perform a parameter estimation through optimization to set the values of the model coefficients that better fit the observed values and describe the spreading. Results confirm that the proposed model fit the data very accurately and we report also some interesting predictions. At the end possible ideas are proposed for further studies.