Prior-predictive value from fast growth simulations

Abstract

Building on a variant of the Jarzynski equation we propose a new method to numerically determine the prior-predictive value in a Bayesian inference problem. The method generalizes thermodynamic integration and is not hampered by equilibration problems. We demonstrate its operation by applying it to two simple examples and elucidate its performance. In the case of multi-modal posterior distributions the performance is superior to thermodynamic integration.

8 pages, 11 figures

References in corpus (6)

• Optimal finite-time processes in stochastic thermodynamics
• Rare events and the convergence of exponentially averaged work values
• Distribution of work in isothermal non-equilibrium processes
• Computing the optimal protocol for finite-time processes in stochastic thermodynamics
• On the efficiency of path sampling methods for the calculation of free energies from non-equilibrium simulations
• Reversibility, heat dissipation and the importance of the thermal environment in stochastic models of nonequilibrium steady states