Towards a new generation of parton densities with deep learning models

Abstract

We present a new regression model for the determination of parton distribution functions (PDF) using techniques inspired from deep learning projects. In the context of the NNPDF methodology, we implement a new efficient computing framework based on graph generated models for PDF parametrization and gradient descent optimization. The best model configuration is derived from a robust cross-validation mechanism through a hyperparametrization tune procedure. We show that results provided by this new framework outperforms the current state-of-the-art PDF fitting methodology in terms of best model selection and computational resources usage.

9 pages, 10 figures

References in corpus (1)

• Mapping the sensitivity of hadronic experiments to nucleon structure

Cited by in corpus (16)

• The Path to Proton Structure at One-Percent Accuracy
• Parton Distributions in Nucleons and Nuclei
• Machine Learning and LHC Event Generation
• nNNPDF2.0: Quark Flavor Separation in Nuclei from LHC Data
• A Neural Resampler for Monte Carlo Reweighting with Preserved Uncertainties
• A guide for deploying Deep Learning in LHC searches: How to achieve optimality and account for uncertainty
• Bayesian Approach to Inverse Problems: an Application to NNPDF Closure Testing
• Determining the proton content with a quantum computer
• Exploring anomalous couplings in Higgs boson pair production through shape analysis
• Parton distributions need representative sampling
• PDFFlow: parton distribution functions on GPU
• Unbinned multivariate observables for global SMEFT analyses from machine learning
• Compressing PDF sets using generative adversarial networks
• A data-based parametrization of parton distribution functions
• Extraction of the color dipole amplitude with physics-informed neural networks
• A new generation of simultaneous fits to LHC data using deep learning