Learning to Simulate High Energy Particle Collisions from Unlabeled Data

Abstract

In many scientific fields which rely on statistical inference, simulations are often used to map from theoretical models to experimental data, allowing scientists to test model predictions against experimental results. Experimental data is often reconstructed from indirect measurements causing the aggregate transformation from theoretical models to experimental data to be poorly-described analytically. Instead, numerical simulations are used at great computational cost. We introduce Optimal-Transport-based Unfolding and Simulation (OTUS), a fast simulator based on unsupervised machine-learning that is capable of predicting experimental data from theoretical models. Without the aid of current simulation information, OTUS trains a probabilistic autoencoder to transform directly between theoretical models and experimental data. Identifying the probabilistic autoencoder's latent space with the space of theoretical models causes the decoder network to become a fast, predictive simulator with the potential to replace current, computationally-costly simulators. Here, we provide proof-of-principle results on two particle physics examples, $Z$-boson and top-quark decays, but stress that OTUS can be widely applied to other fields.

Accepted by Scientific Reports; Changes: Updated title and abstract, rearranged order of sections, added section 4.2, Figure 2, supplementary ablation study, and supplementary figures 2-4; 32 pages, 12 figures, 4 tables

References in corpus (14)

• Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
• Observation of a new particle in the search for the Standard Model Higgs boson with the ATLAS detector at the LHC
• Observation of a new boson at a mass of 125 GeV with the CMS experiment at the LHC
• PYTHIA 6.4 Physics and Manual
• Machine learning for molecular simulation
• From optimal transport to generative modeling: the VEGAN cookbook
• Lorentz Group Equivariant Neural Network for Particle Physics
• Permutationless Many-Jet Event Reconstruction with Symmetry Preserving Attention Networks
• Conditional Variational Autoencoder for Neural Machine Translation
• Linearized Optimal Transport for Collider Events
• Topological Obstructions to Autoencoding
• SARM: Sparse Autoregressive Model for Scalable Generation of Sparse Images in Particle Physics
• Lightning-Fast Gravitational Wave Parameter Inference through Neural Amortization
• Top Quark Mass Measurements in ATLAS and CMS

Cited by in corpus (27)

• Machine Learning and LHC Event Generation
• Event Generators for High-Energy Physics Experiments
• PC-JeDi: Diffusion for Particle Cloud Generation in High Energy Physics
• Calomplification -- The Power of Generative Calorimeter Models
• PC-Droid: Faster diffusion and improved quality for particle cloud generation
• Deep Generative Models for Detector Signature Simulation: A Taxonomic Review
• Deducing Neutron Star Equation of State Parameters Directly From Telescope Spectra with Uncertainty-Aware Machine Learning
• Improving Generative Model-based Unfolding with Schrödinger Bridges
• The Landscape of Unfolding with Machine Learning
• A Method to Simultaneously Facilitate All Jet Physics Tasks
• A simultaneous unbinned differential cross section measurement of twenty-four $Z$+jets kinematic observables with the ATLAS detector
• A Holistic Approach to Predicting Top Quark Kinematic Properties with the Covariant Particle Transformer
• Unbinned Profiled Unfolding
• Full Event Particle-Level Unfolding with Variable-Length Latent Variational Diffusion
• Flows for Flows: Morphing one Dataset into another with Maximum Likelihood Estimation
• CaloChallenge 2022: A Community Challenge for Fast Calorimeter Simulation
• PIPPIN: Generating variable length full events from partons
• Unsupervised and lightly supervised learning in particle physics
• Moment Unfolding
• TURBO: The Swiss Knife of Auto-Encoders
• Advancing Set-Conditional Set Generation: Diffusion Models for Fast Simulation of Reconstructed Particles
• Unifying Simulation and Inference with Normalizing Flows
• Generative Unfolding with Distribution Mapping
• Machine Learning-Assisted Unfolding for Neutrino Cross-section Measurements with the OmniFold Technique
• How to Unfold Top Decays
• Tools for Unbinned Unfolding
• Trials Factor for Semi-Supervised NN Classifiers in Searches for Narrow Resonances at the LHC