Frequentist Uncertainties on Neural Density Ratios with wifi Ensembles

2025/05/30 by Benevedes, Sean, Thaler, Jesse · 2 citations
#Data Analysis #FOS: Physical sciences #High Energy Physics - Experiment (hep-ex) #High Energy Physics - Phenomenology (hep-ph) #Statistics and Probability (physics.data-an)

paper · doi:10.48550/arxiv.2506.00113

Abstract

We introduce wifi ensembles as a novel framework to obtain asymptotic frequentist uncertainties on density ratios, with a particular focus on neural ratio estimation in the context of high-energy physics. When the density ratio of interest is a likelihood ratio conditioned on parameters, wifi ensembles can be used to perform simulation-based inference on those parameters. After training the basis functions f_i(x), uncertainties on the weights w_i can be straightforwardly propagated to the estimated parameters without requiring extraneous bootstraps. To demonstrate this approach, we present an application in quantum chromodynamics at the Large Hadron Collider, using wifi ensembles to estimate the likelihood ratio between generated quark and gluon jets. We use this learned likelihood ratio to estimate the quark fraction in a synthetic mixed quark/gluon sample, showing that the resultant uncertainties empirically satisfy the desired coverage properties.

Citations

• Stabilizing Neural Likelihood Ratio Estimation
• Isolating Unisolated Upsilons with Anomaly Detection in CMS Open Data
• Model-agnostic search for dijet resonances with anomalous jet substructure in proton-proton collisions at √(s) = 13 TeV
• An implementation of neural simulation-based inference for parameter estimation in ATLAS
• CaloChallenge 2022: A Community Challenge for Fast Calorimeter Simulation
• Anomaly Detection in Collider Physics via Factorized Observables
• A Primer on Bayesian Neural Networks: Review and Debates
• End-To-End Latent Variational Diffusion Models for Inverse Problems in High Energy Physics
• EPiC-GAN: Equivariant Point Cloud Generation for Particle Jets
• An unfolding method based on conditional Invertible Neural Networks (cINN) using iterative training
• A comprehensive guide to the physics and usage of PYTHIA 8.3
• A comprehensive guide to the physics and usage of PYTHIA 8.3
• A Trust Crisis In Simulation-Based Inference? Your Posterior Approximations Can Be Unfaithful
• A Living Review of Machine Learning for Particle Physics
• Precision physics with inclusive QCD processes
• Neural Empirical Bayes: Source Distribution Estimation and its Applications to Simulation-Based Inference
• Generative Networks for LHC events
• Learning Multivariate New Physics
• SciPy 1.0: fundamental algorithms for scientific computing in Python
• α_s(2019): Precision measurements of the QCD coupling
• LHC analysis-specific datasets with Generative Adversarial Networks
• Event Generation and Statistical Sampling for Physics with Deep\n Generative Models and a Density Information Buffer
• Determining the strong coupling: status and challenges
• Loss Surfaces, Mode Connectivity, and Fast Ensembling of DNNs
• The strong coupling: a theoretical perspective
• The Top Mass in Hadronic Collisions
• Snapshot Ensembles: Train 1, get M for free
• Simple and Scalable Predictive Uncertainty Estimation using Deep Ensembles
• Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning
• Approximating Likelihood Ratios with Calibrated Discriminative Classifiers
• Weight Uncertainty in Neural Networks
• Adam: A Method for Stochastic Optimization
• Understanding Random Forests: From Theory to Practice
• Pseudo Maximum Likelihood Estimation: Theory and Applications
• The Large-Sample Distribution of the Likelihood Ratio for Testing Composite Hypotheses

Cited by

• Higgs Signal Strength Estimation with Machine Learning under Systematic Uncertainties
• Observable Optimization for Precision Theory: Machine Learning Energy Correlators

Related

• Amplitude Uncertainties Everywhere All at Once
• Extraction of lattice QCD spectral densities from an ensemble of trained machines
• Loop Amplitudes from Precision Networks
• A Kernel Framework to Quantify a Model's Local Predictive Uncertainty under Data Distributional Shifts
• Schrodinger Neural Network and Uncertainty Quantification: Quantum Machine