RT info:eu-repo/semantics/bachelorThesis T1 Machine learning and deep learning in solar physics A1 Zurita Martel, Yazmina K1 Machine Learning K1 Deep Learning K1 Solar AB There exists a growing interest in the application of Machine Learning methods to Astrophysics,as they have proven to be very useful in the past. Following this line of work and motivatedby a reduction in computation time, we explored the use of an artificial neural network (ANN)in the synthesis of Stokes profiles through the direct mapping of solar atmosphere magnitudesand profiles. First, some basic concepts regarding Machine Learning and the problem at handare presented, followed by a complete description of the ANN training process. Finally, a Stokesprofile inversion (which requires a great number of synthesis) is carried out using an optimizationalgorithm to recover the solar atmosphere magnitudes. YR 2020 FD 2020 LK http://riull.ull.es/xmlui/handle/915/20669 UL http://riull.ull.es/xmlui/handle/915/20669 LA es DS Repositorio institucional de la Universidad de La Laguna RD 10-nov-2024