(MC)**3 -- a Multi-Channel Markov Chain Monte Carlo algorithm for phase-space sampling
High Energy Physics - Phenomenology
2015-06-12 v3 Data Analysis, Statistics and Probability
Abstract
A new Monte Carlo algorithm for phase-space sampling, named (MC)**3, is presented. It is based on Markov Chain Monte Carlo techniques but at the same time incorporates prior knowledge about the target distribution in the form of suitable phase-space mappings from a corresponding Multi-Channel importance sampling Monte Carlo. The combined approach inherits the benefits of both techniques while typical drawbacks of either solution get ameliorated.
Cite
@article{arxiv.1404.4328,
title = {(MC)**3 -- a Multi-Channel Markov Chain Monte Carlo algorithm for phase-space sampling},
author = {Kevin Kroeninger and Steffen Schumann and Benjamin Willenberg},
journal= {arXiv preprint arXiv:1404.4328},
year = {2015}
}
Comments
version accepted for publication in CPC, largely extended examples section