Christian Duque Arribas
Universidad Complutense de Madrid, Spain

Abstract
Traditional linear regression models, often solved using ordinary least squares (OLS), frequently overlook measurement uncertainties, leading to potential biases in parameter estimation. This seminar introduces a Bayesian approach to linear regression using Markov Chain Monte Carlo (MCMC) techniques via Stan, showcasing its advantages in handling real-world data uncertainties. As an example, we will analyze the relationship between the mass of the central black hole and the velocity dispersion of the host galaxy, performing Bayesian linear regression both with and without measurement errors. We will explore how Stan's tools make the implementation of Bayesian regression accessible, enabling attendees to apply this methodology in their own research.

2024 November 13, 13:30

IA/U.Porto
Centro de Astrofísica da Universidade do Porto (Classroom)
Rua das Estrelas, 4150-762 Porto