Speaker
Dr
Sabrina Guastavino
(MIDA, Dipartimento di Matematica, Università di Genova)
Description
Solar flare forecasting can be realized by means of either the numerical solution of model equations or the analysis of magnetic data by means of artificial intelligence techniques. Here we consider this second approach and discuss how neural network optimization and forecasting assessment by means of skill scores are deeply intertwined. A special focus is also given to the interpretation of data descriptors in terms of their impact on the prediction effectiveness. Applications will be concerned with magnetograms recorded by SDO/HMI, in the case of both point-in-time images and time series of magnetogram frames.
Primary authors
michele piana
(MIDA, dipartimento di matematica, università di genova)
Dr
Sabrina Guastavino
(MIDA, Dipartimento di Matematica, Università di Genova)
Prof.
Federico Benvenuto
(MIDA, Dipartimento di Matematica, Università di Genova)
Dr
Anna Maria Massone
(Dipartimento di Matematica, Universit`a di Genova, via Dodecaneso 35 16146 Genova, Italy)
Dr
Francesco Marchetti
(Dipartimento di Matematica, Università di Padova)
Prof.
Cristina Campi
(MIDA, DIpartimento di Matematica, Università di Genova)
