Stellar Population Photometric Synthesis with AI of S-PLUS galaxies

We trained a Neural Network that can obtain selected STARLIGHT parameters directly from S-PLUS photometry. The training set consisted of over 55 thousand galaxies with their stellar population parameters obtained from a STARLIGHT application by Cid Fernandes et al. (2005). These galaxies were crossmatched with the S-PLUS iDR 3 database, thus, recovering the photometry for the 12 band filters for 55803 objects. We also considered the spectroscopic redshift for each object which was obtained from the SDSS. Finally, we trained a fully connected Neural Network with the 12-band photometry + redshift as features, and targeted some of the STARLIGHT parameters, such as stellar mass and mean stellar age. The model performed very well for some parameters, for example, the stellar mass, with an error of 0.23 dex. In the future, we aim to apply the model to all S-PLUS galaxies, obtaining never-before-seen photometric synthesis for most objects in the catalogue.