In the machine learning context, a model is trained by minimizing a certain loss function. However, the assessment of the classification performances is achieved by considering different skill scores, which are usually chosen according to the specific application. In this talk, a new class of score-oriented loss functions is presented and analyzed. Then, we discuss the employment of such losses in the optimization of scores that are widely adopted in the context of space weather forecasting, focusing in particular on solar flares classification tasks.
Il seminario verrà svolto per via telematica, verrà registrato e il video verrà caricato su questa pagina per permettere la visione ad un pubblico più ampio. Partecipando all'evento con videocamera o microfono attivi si dà il consenso ad essere registrati. Informativa sulla privacy
Video
