Anomalies and Breakpoint Detection For A Dataset of Firefighters' Operations During The Covid-19 Period In France

Firefighters are exposed to many hazards. The main objective of this study is to apply machine learning techniques to tailor the need for firemen operations to their demands. This strategy enables fire departments to organize their resources, which leads to a reduction of human, material and financial requirements. This work focuses on predicting the number of firefighters' interventions during the sensitive period of the global pandemic COVID-19. Experiments applied to a dataset from 2016 to 2021 provided by the Fire and Rescue Department, SDIS 25, in the region Doubs-France have shown an accurate prediction and revealed the existence of a turning point in August 2020 due to an increase in coronavirus cases in France.