The Simplified Likelihood Framework
Abstract
We discuss the simplified likelihood framework as a systematic approximation scheme for experimental likelihoods such as those originating from LHC experiments. We develop the simplified likelihood from the Central Limit Theorem keeping the next-to-leading term in the large expansion to correctly account for asymmetries. Moreover, we present an efficient method to compute the parameters of the simplified likelihood from Monte Carlo simulations. The approach is validated using a realistic LHC-like analysis, and the limits of the approximation are explored. Finally, we discuss how the simplified likelihood data can be conveniently released in the HepData error source format and automatically built from it, making this framework a convenient tool to transmit realistic experimental likelihoods to the community.
Keywords
Cite
@article{arxiv.1809.05548,
title = {The Simplified Likelihood Framework},
author = {Andy Buckley and Matthew Citron and Sylvain Fichet and Sabine Kraml and Wolfgang Waltenberger and Nicholas Wardle},
journal= {arXiv preprint arXiv:1809.05548},
year = {2019}
}
Comments
24 pages, 6 figures. v2: Minor changes, references added, matches JHEP version. Reference code and demo available at https://gitlab.cern.ch/SimplifiedLikelihood/SLtools