Thinking outside the ROCs: Designing Decorrelated Taggers (DDT) for jet substructure

By J. Dolen, Philip C. Harris, S. Marzani, Salvatore Rappoccio, Nhan Tran. 2016

146 citations recorded by the source corpus.

A bstractWe explore the scale-dependence and correlations of jet substructure observables to improve upon existing techniques in the identification of highly Lorentz-boosted objects. Modified observables are designed to remove correlations from existing theoretically well-understood observables, providing practical advantages for experimental measurements and searches for new phenomena. We study such observables in W jet tagging and provide recommendations for observables based on considerations beyond signal and background efficiencies.

Subfields: Aerospace Engineering, Artificial Intelligence, Nuclear and High Energy Physics

View Paper: https://www.semanticscholar.org/paper/4e895d15b7120a21292a35bee8369fd4f2905d7f

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Aerospace Engineering

1. G. Salam
2. G. Soyez
3. E. Re
4. Cms Collaboration
5. Michael Spannowsky
6. Matteo Cacciari

In Artificial Intelligence

1. Cms Collaboration
2. G. Soyez
3. Atlas Collaboration
4. A. Larkoski
5. Pierre Baldi
6. Peter Sadowski
7. E. Re
8. S. Marzani

In Nuclear and High Energy Physics

1. Cms Collaboration
2. Atlas Collaboration
3. Gudrun Heinrich
4. A. Larkoski
5. M. Schönherr
6. I. Moult
7. Benjamin Nachman
8. S. Höche

Citations in the corpus

• Jet substructure at the Large Hadron Collider (2018, methodology): “To ensure a flat QCD background rejection of 26% over the considered mass and pT range, a decorrelation procedure [167] is applied to N 2 .”
• Les Houches 2017: Physics at TeV Colliders Standard Model Working Group Report (2017, methodology): “3 Towards Improved Performance and Robustness for ATLAS and CMS The strategies used by ATLAS and CMS, namely trimmed D2 [780, 787] and SoftDropped N2 [796] with DDT [800] are specific examples of the broader approaches to two-prong tagging discussed above, and therefore our study allows us to gain i”
• Looking Inside Jets (2019, methodology): “The idea behind the DDT procedure [155] is to instead substitute the cut on N -subjettiness, with a cut on a suitable combination of τ21 and of a function of the ρmMDT = m 2 mMDT/(pt,mMDT) (2).”
• Identification of heavy, energetic, hadronically decaying particles using machine-learning techniques (2020, background): “This understanding benefits from the theoretical progress made in jet substructure studies [3], which can result in reduced systematic uncertainties [15].”
• Jet substructure at the Large Hadron Collider: A review of recent advances in theory and machine learning (2017, background): “decorrelated taggers [506] or convolved substructure [507] manually decorrelate prong-tagging features (such as D2) with the jet mass so that using the former to tag boosted bosons and the latter to localize potential signals is amenable to side-band approaches.”
• Jet Substructure Classification in High-Energy Physics with Deep Neural Networks (2016, background): “The power of the networks described here is limited by the accuracy of these models, and expert features may be more robust to variation among the several existing simulation models [46].”
• The Lund jet plane (2018, background): “eup-mitigation scheme being used. Various potential approaches to address this were highlighted in section3.4. { 24 { useful also in terms of the construction of high-performance decorrelated taggers [104]. There is a potential for a range of other applications, including top-tagging, quark{ gluon d”