Bayesian model comparison in cosmology

TL;DR

This paper discusses a modified Bayesian model comparison approach suitable for cosmology, where traditional methods are hindered by the lack of well-motivated priors, by using data separation to estimate necessary distributions.

Contribution

It introduces a practical method for applying Bayesian model comparison in cosmology despite the absence of well-defined priors, leveraging data separation techniques.

Findings

Enables Bayesian model comparison with separable data sets

Provides a workaround for prior-less models in cosmology

Facilitates model evaluation where traditional Bayesian methods fail

Abstract

The standard Bayesian model formalism comparison cannot be applied to most cosmological models as they lack well-motivated parameter priors. However, if the data-set being used is separable then it is possible to use some of the data to obtain the necessary parameter distributions, the rest of the data being retained for model comparison. While such methods are not fully prescriptive, they provide a route to applying Bayesian model comparison in cosmological situations where it could not otherwise be used.