N-dimensional maximum-entropy tomography via particle sampling

TL;DR

This paper introduces a modified maximum-entropy tomography algorithm that employs particle sampling and MCMC techniques to efficiently reconstruct six-dimensional phase space data, demonstrating successful convergence on synthetic and real datasets.

Contribution

The paper presents a novel six-dimensional phase space tomography method using particle sampling and low-dimensional density estimation within the maximum-entropy framework.

Findings

Successful convergence on synthetic data

Effective application to measured data

Enhanced computational efficiency

Abstract

We propose a modified maximum-entropy (MENT) algorithm for six-dimensional phase space tomography. The algorithm uses particle sampling and low-dimensional density estimation to approximate large sets of high-dimensional integrals in the original MENT formulation. We implement this approach using Markov Chain Monte Carlo (MCMC) sampling techniques and demonstrate convergence of six-dimensional MENT on both synthetic and measured data.