Computational Intelligence for Stellar Magnetic Fields Parameter Determination

In this work we present the preliminary results for the automatic determination of the mean longitudinal magnetic field of polarized spectra through the analysis of spectropolarimetric observations. In order to achieve this goal, we first developed an artificial database encompassing a set of different stellar spectra each one defined by a set of free parameters. Then we used supervised learning for artificial neural networks, a computational intelligence paradigm, in order to determine this important parameter.