Speaker
Jean Pierre Cifuentes Salazar
(Estudiante de maestría (UNAL))
Description
This project, carried out within the ATLAS collaboration, focused on jet substructure studies using the Lund jet plane. Lund plane variables were implemented and used to train graph-based machine learning models for tagging hadronically decaying $W$ bosons and $top$ quarks. The performance of these taggers was systematically evaluated using different Monte Carlo event generators, with dedicated studies on the impact of kt-based cuts. These results contribute to understanding the robustness of jet tagging methods under variations in parton shower and hadronization modeling.
Author
Jean Pierre Cifuentes Salazar
(Estudiante de maestría (UNAL))
Co-authors
Carlos Sandoval Usme
(Universidad Nacional de Colombia)
Rafael Andrei Vinasco Soler
(Universidad Nacional de Colombia (CO))
Reina Coromoto Camacho Toro
(LPNHE-Paris CNRS/IN2P3)
