Antonio Carlos Oliveira Da Silva

arXiv2024

Multiplicity Based Background Subtraction for Jets in Heavy Ion Collisions

Jet measurements in heavy ion collisions at low jet momentum can provide constraints on the properties of the quark gluon plasma but are overwhelmed by a significant, fluctuating background. We build upon our previous work which demonstrated the ability of the jet multiplicity method to extend jet measurements into the domain of low jet momentum [1, Mengel:2023]. We extend this method to a wide range of jet resolution parameters. We investigate the over-complexity of non-interpretable machine learning used to tackle the problem of jet background subtraction through network optimization. Finally, we show that the resulting shallow neural network is able to learn the underlying relationship between jet multiplicity and background fluctuations, with a lesser complexity, reinforcing the utility of interpretable methods.

hep-ex↗

arXiv2022

Model studies of fluctuations in the background for jets in heavy ion collisions

Jets produced in high energy heavy ion collisions are quenched by the quark gluon plasma. Measurements of these jets are influenced by the methods used to suppress and subtract the large, fluctuating background and the assumptions inherent in these methods. We compare the measurements of the background in Pb+Pb collisions at $\sqrt{s_{NN}}$ = 2.76 TeV by the ALICE Collaboration $\underline{(B. Abelev \textit{et al.}, J. High Energy Phys. \mathbf{2012}, 053)}$ to calculations in TennGen (a data-driven random background generator) and PYTHIA Angantyr. A detailed understanding of the width of these fluctuations is important for reducing uncertainties due to unfolding and extending measurements to lower momenta and larger resolution parameters. The standard deviation of the energy in random cones in TennGen is approximately in agreement with the form predicted in the ALICE paper, with deviations of 1--6%. The standard deviation of energy in random cones in Angantyr exceeds the same predictions by approximately 13%. Deviations in both models can be explained by the assumption that the single-particle $d^2N/dy$ $dp_T$ is a gamma distribution in the derivation of the prediction, whereas the model uses a different distribution. This indicates that model comparisons are potentially sensitive to the treatment of the background. We demonstrate that unfolding methods used to remove background fluctuations from jets can affect the comparisons between models and data, $\textit{even in the absence of detector effects}$. Our findings suggest the need to more carefully consider methods for comparing simulations and data.

hep-ph↗