Speaker
Description
Four-top-quark production at the LHC is a rare Standard Model process and an important probe of top-quark interactions and possible new physics. The fully hadronic decay channel offers a large branching fraction, but its reconstruction is challenging due to the high jet multiplicity and severe combinatorial ambiguity. In this study, we investigate the reconstruction of fully hadronic (t\bar{t}t\bar{t}) events by assigning the three partons from each hadronically decaying top quark to the corresponding reconstructed jets. We compare two improved machine-learning-based assignment algorithms, EVENet and SPANet, using both stand-alone Monte Carlo samples and CMS Open Data. The performance is evaluated in terms of correct jet–parton assignment and top-quark reconstruction efficiency. In addition, we study the application of a Variable Radius jet-clustering algorithm to this final state. Compared with fixed-radius jet clustering, the Variable Radius approach increases the number of top quark which all three daughter partons can be matched to distinct reconstructed jets, while maintaining a comparable reconstruction efficiency. These results demonstrate the potential of combining adaptive jet reconstruction with advanced assignment networks for fully hadronic four-top analyses.
| 请选择分会 | TeV物理和超出标准模型新物理 |
|---|
