Tag: Wildfires

Researchers from VSB-Technical University of Ostrava, Czech Republic, along with teams from Argonne and Oak Ridge National Laboratories, have developed a machine learning model that uses satellite images to identify wildfires in remote, boreal forests like Alaska. The model employs Google Earth Engine API and PERMON and PETSc frameworks to train classification models of a maximal-margin type. The researchers are also extending the PERMON toolbox for training machine learning models, discussing adaptations of solvers for deep learning techniques, and developing machine learning approaches for processing and analyzing seismic events, supported by the Strategy AV21 project.