Leveraging universality of jet taggers through transfer learning
arXiv:2203.06210 · doi:10.1140/epjc/s10052-022-10469-9
Abstract
A significant challenge in the tagging of boosted objects via machine-learning technology is the prohibitive computational cost associated with training sophisticated models. Nevertheless, the universality of QCD suggests that a large amount of the information learnt in the training is common to different physical signals and experimental setups. In this article, we explore the use of transfer learning techniques to develop fast and data-efficient jet taggers that leverage such universality. We consider the graph neural networks LundNet and ParticleNet, and introduce two prescriptions to transfer an existing tagger into a new signal based either on fine-tuning all the weights of a model or alternatively on freezing a fraction of them. In the case of -boson and top-quark tagging, we find that one can obtain reliable taggers using an order of magnitude less data with a corresponding speed-up of the training process. Moreover, while keeping the size of the training data set fixed, we observe a speed-up of the training by up to a factor of three. This offers a promising avenue to facilitate the use of such tools in collider physics experiments.
arXiv comment10 pages, 2 tables, 5 figures
References in corpus (47)
- The anti-k_t jet clustering algorithm
- An Introduction to PYTHIA 8.2
- Better Jet Clustering Algorithms
- Identification of heavy-flavour jets with the CMS detector in pp collisions at 13 TeV
- ParticleNet: Jet Tagging via Particle Clouds
- Deep Learning and its Application to LHC Physics
- Jet Substructure at the Large Hadron Collider: A Review of Recent Advances in Theory and Machine Learning
- NNLOPS simulation of Higgs boson production
- Deep learning in color: towards automated quark/gluon jet discrimination
- Jet-Images -- Deep Learning Edition
- Deep-learning Top Taggers or The End of QCD?
- Anomaly Detection for Resonant New Physics with Machine Learning
- Learning New Physics from a Machine
- Extending the Bump Hunt with Machine Learning
- Deep-learned Top Tagging with a Lorentz Layer
- Jet Flavor Classification in High-Energy Physics with Deep Neural Networks
- Parton showers beyond leading logarithmic accuracy
- QCD or What?
- Identification of heavy, energetic, hadronically decaying particles using machine-learning techniques
- Simulation Assisted Likelihood-free Anomaly Detection
- Drell-Yan lepton pair production at NNLO QCD with parton showers
- Parton Shower Uncertainties in Jet Substructure Analyses with Deep Neural Networks
- Energy flow polynomials: A complete linear basis for jet substructure
- Pulling Out All the Tops with Computer Vision and Deep Learning
- The Lund Jet Plane
- Guiding New Physics Searches with Unsupervised Learning
- JEDI-net: a jet identification algorithm based on interaction networks
- QCD-Aware Recursive Neural Networks for Jet Physics
- Logarithmic accuracy of parton showers: a fixed-order study
- Matching Fully Differential NNLO Calculations and Parton Showers
- Boosting with Machine Learning
- MiNNLO: A new method to match NNLO QCD to parton showers
- Implementing NLO DGLAP evolution in Parton Showers
- Recursive Neural Networks in Quark/Gluon Tagging
- MiNNLO: Optimizing hadronic processes
- Interaction networks for the identification of boosted decays
- Measurement of the Lund jet plane using charged particles in 13 TeV proton-proton collisions with the ATLAS detector
- Summations of large logarithms by parton showers
- Interpretable Deep Learning for Two-Prong Jet Classification with Jet Spectra
- Mass Unspecific Supervised Tagging (MUST) for boosted jets
- Automating the Construction of Jet Observables with Machine Learning
- Matching NNLO to parton shower using NLL colour-singlet transverse momentum resummation in GENEVA
- Towards Machine Learning Analytics for Jet Substructure
- Calculating the primary Lund Jet Plane density
- Higgs boson tagging with the Lund jet plane
- Spectral Analysis of Jet Substructure with Neural Networks: Boosted Higgs Case
- Matching N3LO QCD calculations to parton showers