English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

The particle-flow (PF) algorithm, which infers particles based on tracks andcalorimeter clusters, is of central importance to event reconstruction in theCMS experiment at the CERN LHC, and has been a focus of development in light ofplanned Phase-2 running conditions with an increased pileup and detectorgranularity. In recent years, the machine learned particle-flow (MLPF)algorithm, a graph neural network that performs PF reconstruction, has beenexplored in CMS, with the possible advantages of directly optimizing for thephysical quantities of interest, being highly reconfigurable to new conditions,and being a natural fit for deployment to heterogeneous accelerators. Wediscuss progress in CMS towards an improved implementation of the MLPFreconstruction, now optimized using generator/simulation-level particleinformation as the target for the first time. This paves the way to potentiallyimproving the detector response in terms of physical quantities of interest. Wedescribe the simulation-based training target, progress and studies onevent-based loss terms, details on the model hyperparameter tuning, as well asphysics validation with respect to the current PF algorithm in terms ofhigh-level physical quantities such as the jet and missing transverse momentumresolutions. We find that the MLPF algorithm, trained on a generator/simulatorlevel particle information for the first time, results in broadly compatibleparticle and jet reconstruction performance with the baseline PF, setting thestage for improving the physics performance by additional training statisticsand model tuning.

Progress towards an improved particle flow algorithm at CMS with machine learning