Speaker
Andrea Sciandra
Description
The ParticleNet tagger is a graph neural network devoted to the tagging of jets from the hadronization of multiple flavors. Its impressive and unprecedented tagging performance allows for accessing rare and challenging hadronic final states. This study shows the fast-simulation-based characterization of the ParticleNet performance evolution as a function of the IDEA vertex detector single-hit resolution, material radiation length and number of layers. Furthermore, an attempt to study impacts in physics applications such as the Z(qq)H and Z(inv)H final states will be shown.
Primary authors
Andrea Sciandra
George Iakovidis
(Brookhaven National Laboratory)
Viviana Cavaliere
(Brookhaven National Laboratory)
