Fully Bayesian Forecasts with Evidence Networks

TL;DR

This paper introduces a new simulation-based method for efficient Bayesian model comparison forecasts that marginalize over uncertainties, improving the design of experiments and theoretical research.

Contribution

It presents a novel, computationally feasible approach for Bayesian forecasts that avoids restrictive assumptions and enhances model comparison accuracy.

Findings

Enables rapid Bayesian model comparison forecasts

Marginalizes over uncertain parameters and noise

Improves experiment design and theoretical efforts

Abstract

Sensitivity forecasts inform the design of experiments and the direction of theoretical efforts. To arrive at representative results, Bayesian forecasts should marginalize their conclusions over uncertain parameters and noise realizations rather than picking fiducial values. However, this is typically computationally infeasible with current methods for forecasts of an experiment's ability to distinguish between competing models. We thus propose a novel simulation-based methodology capable of providing expedient and rigorous Bayesian model comparison forecasts without relying on restrictive assumptions.