Dove e quando

speaker: Sabrina Guastavino
affiliation: Dipartimento di Matematica, Università di Genova
date: February 15^th, 2023
venue: Torino (Italy), Dipartimento di Scienze Matematiche G.L. Lagrange, Politecnico di Torino

Some key points in machine/deep learning in solar flare forecasting

abstract: In the last decades machine and deep learning approaches have been obtaining an increasing interest in many forecasting problems in many applications as in space weather.  Solar flares are the main trigger of the explosive events associated to the active Sun and the application of artificial intelligence techniques for their prediction is currently an hot topic in space weather.
In this talk we discuss some crucial aspects for the effectiveness of machine/deep learning in forecasting problems: (1) the optimization of the network weights depends on an appropriate choice of the loss function in the training phase; (2) the validation strategy for the prediction method depends on the way datasets are generated for training and validation and (3) the assessment of the prediction power of the method depends on a suitable choice of the evaluation metrics. We will finally show that these key points are crucial both in feature-based machine learning and image-based deep learning for the problem of solar flare forecasting.

https://www.disma.polito.it/