Determining distances of quasars in S-PLUS using deep learning

• Grant number: 21/08983-0
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Effective date (Start): October 01, 2021
• Effective date (End): December 31, 2022
• Field of knowledge: Physical Sciences and Mathematics - Astronomy - Extragalactic Astrophysics
• Principal Investigator: Claudia Lucia Mendes de Oliveira
• Grantee: Raquel Ruiz Valença
• Host Institution: Instituto de Astronomia, Geofísica e Ciências Atmosféricas (IAG). Universidade de São Paulo (USP). São Paulo , SP, Brazil
• Associated research grant: 19/26492-3 - Science with the Brazilian robotic telescope, AP.ESP

Abstract

The main goals of this research project consist of training and evaluating predictive models for redshifts of quasars using photometric data from the Southern Photometric Local Universe Survey (S-PLUS) in a supervised machine learning context. S-PLUS data from quasars previously confirmed via spectroscopy by the Sloan Digital Sky Survey (SDSS) will be used for model training. In this project, we will focus on the study of deep learning models, specifically Bayesian neural network models. Metrics such as precision, bias, and outlier fraction will be used to assess the performance of the models. The best model obtained will be applied to quasar candidates that were previously photometrically classified.(AU)