Machine learning-based event generator for electron-proton scattering

TL;DR

This paper introduces a machine learning-based event generator for electron-proton scattering using GANs, capable of producing realistic vertex-level events without relying on theoretical assumptions, thus reducing bias in physical measurements.

Contribution

The paper develops a GAN-based framework for event generation and detector simulation, providing a novel, fast, and unbiased method validated on simulated data with uncertainty quantification.

Findings

Successfully generates realistic vertex-level events

Validates framework on simulated deep-inelastic scattering data

Quantifies uncertainties using bootstrap techniques

Abstract

We present a new machine learning-based Monte Carlo event generator using generative adversarial networks (GANs) that can be trained with calibrated detector simulations to construct a vertex-level event generator free of theoretical assumptions about femtometer scale physics. Our framework includes a GAN-based detector folding as a fast-surrogate model that mimics detector simulators. The framework is tested and validated on simulated inclusive deep-inelastic scattering data along with existing parametrizations for detector simulation, with uncertainty quantification based on a statistical bootstrapping technique. Our results provide for the first time a realistic proof-of-concept to mitigate theory bias in inferring vertex-level event distributions needed to reconstruct physical observables.