Machine Learning for Particle Flow Reconstruction at CMS
Problem
Reconstructing stable particles in the CMS detector from combined tracks and calorimeter clusters (particle flow) to improve jets and missing transverse energy.
Why it matters
Global event reconstruction quality directly affects CMS physics measurements at the LHC.
What the AI did
A graph neural network reconstructs particle candidates from the full event's tracks and calorimeter clusters, studied as a GPU-friendly evolution of particle flow and validated inside the CMS software for jets and missing transverse energy.
Results
Validation in the CMS software framework reported; computational performance scales approximately linearly; no numeric results in abstract.
Impact
Shows an ML-based particle flow algorithm is feasible and validated within CMS offline reconstruction.
Record
| Venue | Journal of Physics Conference Series |
|---|---|
| Published | 2023 |
| Work type | conference-paper |
| Authors | Joosep Pata, Javier Mauricio Duarte, Farouk Mokhtar, Eric Wulff, J. Yoo, Jean-Roch Vlimant, M. Pierini, Maria Girone |
| Citations | 19 (OpenAlex, at import time) |
| References | 60 |
| Model or method | graph neural network |
| Organization | CMS |
| Primary source | https://doi.org/10.1088/1742-6596/2438/1/012100 |
| Open access | diamond |
| Keywords | graph neural network, particle flow, CMS, jets, missing transverse energy, GPU |
Importance rating
significant — thousands to millions affected
Abstract states global event reconstruction quality strongly affects CMS physics measurements (jets, missing transverse energy) at the LHC; significant for a foundational scientific field, single experiment scope.
More in Physics & Astronomy
| # | Paper | Year | Citations | Importance |
|---|---|---|---|---|
| 1 | Magnetic control of tokamak plasmas through deep reinforcement learning | 2022 | 739 |
7
|
| 2 | Solving the quantum many-body problem with artificial neural networks | 2017 | 2,205 |
7
|
| 3 | Learning high-accuracy error decoding for quantum processors | 2024 | 90 |
7
|
| 4 | Optimizing Quantum Error Correction Codes with Reinforcement Learning | 2019 | 158 |
7
|
| 5 | Machine-learning-assisted correction of correlated qubit errors in a topological code | 2018 | 141 |
7
|