SciPost Submission Page
Two Invertible Networks for the Matrix Element Method
by Anja Butter, Theo Heimel, Till Martini, Sascha Peitzsch, Tilman Plehn
This Submission thread is now published as
Submission summary
| Authors (as registered SciPost users): | Theo Heimel · Tilman Plehn |
| Submission information | |
|---|---|
| Preprint Link: | https://arxiv.org/abs/2210.00019v5 (pdf) |
| Date accepted: | 2023-07-20 |
| Date submitted: | 2023-06-22 09:51 |
| Submitted by: | Heimel, Theo |
| Submitted to: | SciPost Physics |
| Ontological classification | |
|---|---|
| Academic field: | Physics |
| Specialties: |
|
Abstract
The matrix element method is widely considered the ultimate LHC inference tool for small event numbers. We show how a combination of two conditional generative neural networks encodes the QCD radiation and detector effects without any simplifying assumptions, while keeping the computation of likelihoods for individual events numerically efficient. We illustrate our approach for the CP-violating phase of the top Yukawa coupling in associated Higgs and single-top production. Currently, the limiting factor for the precision of our approach is jet combinatorics.
List of changes
We would like to thank the referees for their comments. We uploaded an updated version with the requested changes.
Report by Sebastien Wertz:
1. We changed the first two sentences of the abstract to "The matrix element method is widely considered the ultimate LHC inference tool for small event numbers. We show how a combination of two conditional generative neural networks encodes the QCD radiation and detector effects without any simplifying assumptions, while keeping the computation of likelihoods for individual events numerically efficient."
2. We corrected the typo.
3. Indeed, it might be useful when ISR is included in the leptonic channel, for example to apply a sorting of the reco-level jets that makes the combinatorics easier to extract for the network.
4. We changed the text to "our xhard-integration would become trivial if we could sample".
5. We updated Figure 4 to include 1/p(r).
Anonymous report:
We changed the part of the introduction about earlier uses of ML for the MEM to "A connection between the MEM and modern ML-methods, mentioned in Ref. [67], was demonstrated in Ref. [68], specifically using a deep regression network to evaluate the MEM integral."
[67] HEP Software Foundation, J. Albrecht et al., A Roadmap for HEP Software and Computing R&D for the 2020s
[68] F. Bury and C. Delaere, Matrix element regression with deep neural networks — Breaking the CPU barrier
Published as SciPost Phys. 15, 094 (2023)