Application of AI to the immediate prediction of time series to optimize resource management in epidemics

Given the uncertainty generated by the health crisis, and from the large amount of data that is generated every day on the COVID19 disease, we propose to apply the most modern and reliable techniques for analysis and prediction of spatio-temporal series using / deep-learning / and others paradigms of artificial intelligence. The objective is to immediately develop predictive systems for infections, admissions, ICU patients and deaths by province and autonomous community, together with other magnitudes (logistical, economic, ...) that allow to guide more effectively the action of the Ministry of Health in this and other crises. In addition, a retrospective analysis is proposed that makes it possible to determine a posteriori the start date of each outbreak of the epidemic as well as visualization tools and dashboards that allow the extraction of useful knowledge and facilitate decision-making based on the data.