Speaker
Description
During Run 3, the LHCb experiment operates at an instantaneous luminosity approximately five times higher than in Run 2, leading to a substantial increase in event multiplicity and data volume. Each proton–proton collision produces hundreds of tracks, while only a small fraction corresponds to the signal. To reduce event size without compromising physics performance, LHCb has developed inclusive isolation tools that retain only the information relevant for detailed offline analysis. Among these, the Inclusive Multivariate Isolation (IMI) algorithm is designed to robustly separate signal tracks from high-pileup background across a wide range of decay topologies and kinematic regimes. Deployed in both the High-Level Trigger and offline reconstruction, IMI combines traditional isolation concepts with multivariate techniques, achieving a 45% reduction in total data size while maintaining signal efficiencies above 99%. Its stability and robustness have been validated under real data-taking conditions. Looking ahead to the future LHCb environment at even higher luminosities, new methods, including more sophisticated techniques such as graph neural networks, will be developed for the signal reconstruction of beauty decays. This contribution is based on arXiv:2511.11487, which is accepted for publication on EPJC.
| I read the instructions above | Yes |
|---|