A novel Bayesian approach to the computation of the configurational density of states

TL;DR

This paper introduces a Bayesian method for efficiently computing the density of states in physical systems, adaptable to various ensembles, and capable of producing accurate results with reasonable computational effort.

Contribution

A new Bayesian approach utilizing test functions and Bayes theorem to accurately estimate the density of states across different ensembles.

Findings

Algorithm finds DOS in reasonable time

Results closely match true DOS with suitable test functions

Method is adaptable to various ensemble distributions

Abstract

In this work we develop and implement a novel Bayesian method for computing the DOS of a system. This method is based on the use of a test function with adjustable parameters and we use Bayes theorem to find the best parameters given a certain number of measurements done on the system. This measurements can be done in any ensemble defined by a distribution function. We found that the algorithm can find the DOS in a reasonable amount of time, and that if the test function is suitable enough, the DOS found by the algorithm is very close to the true DOS.