Calculating Bayesian evidence for inflationary models using CONNECT

Abstract

Bayesian evidence is a standard tool used for comparing the ability of different models to fit available data and is used extensively in cosmology. However, since the evidence calculation involves performing an integral of the likelihood function over the entire space of model parameters this can be prohibitively expensive in terms of both CPU and time consumption. For example, in the simplest $Λ$CDM model and using CMB data from the Planck satellite, the dimensionality of the model space is over 30 (typically 6 cosmological parameters and 28 nuisance parameters). Even the simplest possible model requires $\mathcal{O}(10^6)$ calls to an Einstein--Boltzmann solver such as CLASS or CAMB and takes several days. Here we present calculations of Bayesian evidence using the CONNECT framework to calculate cosmological observables. We demonstrate that we can achieve results comparable to those obtained using Einstein--Boltzmann solvers, but at a minute fraction of the computational cost. As a test case, we then go on to compute Bayesian evidence ratios for a selection of slow-roll inflationary models. In the setup presented here, the total computation time is completely dominated by the likelihood function calculation which now becomes the main bottleneck for increasing computation speed.