SciPost Submission Page
Inferring flavor mixtures in multijet events
by Ezequiel Alvarez, Yuling Yao
This Submission thread is now published as
Submission summary
| Authors (as registered SciPost users): | Ezequiel Alvarez |
| Submission information | |
|---|---|
| Preprint Link: | scipost_202409_00020v1 (pdf) |
| Date accepted: | 2024-10-29 |
| Date submitted: | 2024-09-17 19:46 |
| Submitted by: | Alvarez, Ezequiel |
| Submitted to: | SciPost Physics |
| Ontological classification | |
|---|---|
| Academic field: | Physics |
| Specialties: |
|
| Approaches: | Experimental, Computational, Phenomenological |
Abstract
Multijet events with heavy-flavors are of central importance at the LHC since many relevant processes -- such as $t\bar t$, $hh$, $t\bar t h$ and others -- have a preferred branching ratio for this final state. Current techniques for tackling these processes use hard-assignment selections through $b$-tagging working points, and suffer from systematic uncertainties because of the difficulties in Monte Carlo simulations. We develop a flexible Bayesian mixture model approach to simultaneously infer $b$-tagging score distributions and the flavor mixture composition in the dataset. We model multidimensional jet events, and to enhance estimation efficiency, we design structured priors that leverages the continuity and unimodality of the $b$-tagging score distributions. Remarkably, our method eliminates the need for a parametric assumption and is robust against model misspecification -- It works for arbitrarily flexible continuous curves and is better if they are unimodal. We have run a toy inferential process with signal $bbbb$ and backgrounds $bbcc$ and $cccc$, and we find that with a few hundred events we can recover the true mixture fractions of the signal and backgrounds, as well as the true $b$-tagging score distribution curves, despite their arbitrariness and nonparametric shapes. We discuss prospects for taking these findings into a realistic scenario in a physics analysis. The presented results could be a starting point for a different and novel kind of analysis in multijet events, with a scope competitive with current state-of-the-art analyses. We also discuss the possibility of using these results in general cases of signals and backgrounds with approximately known continuous distributions and/or expected unimodality.
Author indications on fulfilling journal expectations
- Provide a novel and synergetic link between different research areas.
- Open a new pathway in an existing or a new research direction, with clear potential for multi-pronged follow-up work
- Detail a groundbreaking theoretical/experimental/computational discovery
- Present a breakthrough on a previously-identified and long-standing research stumbling block
Author comments upon resubmission
Please find enclosed the new version of the manuscript responding to all the points suitably indicated by the Referees. For their convenience, all the changes have been marked in blue.
With kind regards,
Yuling and Ezequiel.
List of changes
All changes required by the Referees are in blue font in the new version of the manuscript
Published as SciPost Phys. Core 7, 076 (2024)
Reports on this Submission
Strengths
1. The paper presents a novel statistical approach for going beyond fixed working-point jet flavour taggers when analyzing multi-component and multi-jet event datasets.
2. The ideas presented in this work have multiple potential generalizations and applications to other similar problems in high energy particle phenomenology.
Weaknesses
1. The methods are tested on toy examples. Consequently, their performance and viability in realistic scenarios is difficult to evaluate.
Report
In their revised version, the authors have addressed all points raised in my first report. Nonetheless the weakness of working only with toy examples remains. Thus I would recommend the publication of the manuscript in SciPost Core.
Recommendation
Accept in alternative Journal (see Report)
Report #2 by Anonymous (Referee 2) on 2024-10-10 (Invited Report)
- Cite as: Anonymous, Report on arXiv:scipost_202409_00020v1, delivered 2024-10-10, doi: 10.21468/SciPost.Report.9889
Strengths
Referees' specific requests on presentation and explanation sufficiently implemented.
Weaknesses
The study remains fundamentally a toy one based around essentially closure tests on uncorrelated distributions, lacking the complexity of realistic events, and the physics use-case for distribution-level (rather than event-level) flavour disambiguation is not clear.
Report
See comments above for specific strengths and weaknesses. I find the presentation and accessibility to a physics audience improved wrt the original, for which thanks. The clarity could still be improved by more explicit, direct and simple explanations, but it is comprehensible and publishable with the main aspects more completely specified than before.
But fundamentally the limitations of the toy study make it hard to see the direct physics application of distribution demixing -- unfolding or background-subtraction, perhaps? It is unrealistic to expect a complete upgrade to a more realistic simulation and physics application in a paper revision, so I am happy to see this published as-is, but in SciPost Physics Core to reflect the degree of novelty and likely impact.
Requested changes
None
Recommendation
Accept in alternative Journal (see Report)
Report
I appreciate the effort the authors put into answering my questions, and I understand that some of my issues have to be postponed to another paper. While I do think that the current version of the paper is not perfect, I would be happy to settle for SciPost Core, and I am looking forward to a more quantitative and detailed follow-up.
Recommendation
Accept in alternative Journal (see Report)