Title: Astrostatistical challenges in the next decade of cosmology
Authors: Roberto Trotta - Imperial College London (United Kingdom) [presenting]
Abstract: Thanks to large and accurate measurements obtained in the last two decades, and to sophisticated statistical analyses, cosmologists have established a cosmological concordance model that reproduces well observations ranging from the relic radiation from the Big Bang to the distribution of galaxies in the sky in the modern Universe. We will introduce and review the observational and theoretical underpinnings of this so-called Lambda-CDM concordance model of cosmology, which strongly points to the existence of both dark matter and dark energy -- whose nature is presently the most important outstanding question of cosmology. We will then present the statistical and data science challenges associated with upcoming large observational data from large telescope and space missions. We will discuss how AI and machine learning will be (and already are) indispensable tools to interpret and analyse data about the universe, together with sophisticated Bayesian modeling. We will make the case that we need to go beyond a ``black box'' approach to inference and model selection, and exploit fully our understanding of the underpinning physics, and our ability to model the data from first principle, including complex selection effects.