Deeply learning deep inelastic scattering kinematics

Diefenthaler, M.; Farhat, Abdullah; Verbytskyi, A.; Xu, Yuesheng

doi:10.1140/epjc/s10052-022-10964-z

Cited by 9 publications

(8 citation statements)

References 53 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Momentum and energy conservation in DIS kinematics provide the ability to calculate x, Q 2 , and y from measurements. Classical methods for their reconstruction differ (see [3,4,9]). We compare our results with methods such as electron (EL), double angle (DA), and Jacquet Blondel (JB).…”

Section: Dismentioning

confidence: 99%

“…We compare our results with methods such as electron (EL), double angle (DA), and Jacquet Blondel (JB). As highlighted in [3,4], the DIS process can be influenced by several factors, such as initial-state and final-state radiation (ISR, FSR). Moreover, higher-order quantum electrodynamics (QED) and quantum chromodynamics (QCD) corrections can also manifest in the process.…”

Section: Dismentioning

confidence: 99%

“…They reduce the computational overhead of using a normalizing flow by allowing Additional effects may arise from, e.g. higher-order QED corrections at the lepton vertex or QCD corrections (see [3,4]).…”

Section: Eluquant Architecturementioning

confidence: 99%

“…Deep inelastic scattering (DIS) has recently benefited from deep learning techniques. An innovative study by Diefenthaler et al [3] employed deep neural networks (DNNs) to infer kinematic variables Q 2 and x of neutral current DIS from traditional reconstruction methods enhanced through correlations revealed in the simulated datasets of the ZEUS experiment. Successively, Arratia et al [4] applied DNNs, capitalizing on full kinematic information of both the scattered electron and the hadronic-final state (HFS), to reconstruct the kinematics of neutralcurrent DIS events, using H1 experiment simulations.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

ELUQuant: event-level uncertainty quantification in deep inelastic scattering

Fanelli,

Giroux

2024

Mach. Learn.: Sci. Technol.

View full text Add to dashboard Cite

We introduce a physics-informed Bayesian Neural Network (BNN) with flow approximated posteriors using multiplicative normalizing flows (MNF) for detailed uncertainty quantification (UQ) at the physics event-level. Our method is capable of identifying both heteroskedastic aleatoric and epistemic uncertainties, providing granular physical insights. Applied to Deep Inelastic Scattering (DIS) events, our model effectively extracts the kinematic variables $x$, $Q^2$, and $y$, matching the performance of recent deep learning regression techniques but with the critical enhancement of event-level UQ. This detailed description of the underlying uncertainty proves invaluable for decision-making, especially in tasks like event filtering. It also allows for the reduction of true inaccuracies without directly accessing the ground truth. A thorough DIS simulation using the H1 detector at HERA indicates possible applications for the future EIC. Additionally, this paves the way for related tasks such as data quality monitoring and anomaly detection. Remarkably, our approach effectively processes large samples at high rates.

show abstract

Section: Dismentioning

confidence: 99%

Section: Dismentioning

confidence: 99%

Section: Eluquant Architecturementioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

ELUQuant: event-level uncertainty quantification in deep inelastic scattering

Fanelli,

Giroux

2024

Mach. Learn.: Sci. Technol.

View full text Add to dashboard Cite

show abstract

“…Using multiple layers, DNNs can effectively discover multi-scale features and representations in input data. Besides its applications in the traditional machine learning field, DNNs have been widely employed in many domains, including physics [12,23], scientific computing [9], and finance [11]. The great successes of DNNs are, to a great extent, due to their mighty expressiveness in representing a function.…”

mentioning

confidence: 99%

Comparison of Contents Design in Mathematical Inquiry in High School Textbooks

2021

School Mathematics Textbooks in China

View full text Add to dashboard Cite

Ahanjideh, Akbari, Fakharan and Trevisan proposed a conjecture on the distribution of the Laplacian eigenvalues of graphs: for any connected graph of order n with diameter d ≥ 2 that is not a path, the number of Laplacian eigenvalues in the interval [n − d + 2, n] is at most n − d. We show that the conjecture is true and provide a family of graphs to show that the bound is best possible. This establishes an interesting relation between the spectral and classical parameters.

show abstract

Machine learning-based jet and event classification at the Electron-Ion Collider with applications to hadron structure and spin physics

Lee

Mulligan

Płoskoń

et al. 2023

J. High Energ. Phys.

View full text Add to dashboard Cite

We explore machine learning-based jet and event identification at the future Electron-Ion Collider (EIC). We study the effectiveness of machine learning-based classifiers at relatively low EIC energies, focusing on (i) identifying the flavor of the jet and (ii) identifying the underlying hard process of the event. We propose applications of our machine learning-based jet identification in the key research areas at the future EIC and current Relativistic Heavy Ion Collider program, including enhancing constraints on (transverse momentum dependent) parton distribution functions, improving experimental access to transverse spin asymmetries, studying photon structure, and quantifying the modification of hadrons and jets in the cold nuclear matter environment in electron-nucleus collisions. We establish first benchmarks and contrast the estimated performance of flavor tagging at the EIC with that at the Large Hadron Collider. We perform studies relevant to aspects of detector design including particle identification, charge information, and minimum transverse momentum capabilities. Additionally, we study the impact of using full event information instead of using only information associated with the identified jet. These methods can be deployed either on suitably accurate Monte Carlo event generators, or, for several applications, directly on experimental data. We provide an outlook for ultimately connecting these machine learning-based methods with first principles calculations in quantum chromodynamics.

show abstract

Deeply learning deep inelastic scattering kinematics

Cited by 9 publications

References 53 publications

ELUQuant: event-level uncertainty quantification in deep inelastic scattering

ELUQuant: event-level uncertainty quantification in deep inelastic scattering

Comparison of Contents Design in Mathematical Inquiry in High School Textbooks

Machine learning-based jet and event classification at the Electron-Ion Collider with applications to hadron structure and spin physics

Contact Info

Product

Resources

About