LHC analysis-specific datasets with Generative Adversarial Networks

TL;DR

This paper explores using GANs to generate analysis-specific high-level features for LHC events, aiming to reduce computational costs in simulations by focusing on relevant distributions.

Contribution

It introduces a GAN-based method tailored for analysis-specific feature generation, including regression-enhanced training and a new performance assessment criterion.

Findings

GANs can generate high-level features with high fidelity.

Regression terms improve GAN training speed and accuracy.

Proposed evaluation metric quantifies generator performance.

Abstract

Using generative adversarial networks (GANs), we investigate the possibility of creating large amounts of analysis-specific simulated LHC events at limited computing cost. This kind of generative model is analysis specific in the sense that it directly generates the high-level features used in the last stage of a given physics analyses, learning the N-dimensional distribution of relevant features in the context of a specific analysis selection. We apply this idea to the generation of muon four-momenta in $Z \to \mu\mu$ events at the LHC. We highlight how use-case specific issues emerge when the distributions of the considered quantities exhibit particular features. We show how substantial performance improvements and convergence speed-up can be obtained by including regression terms in the loss function of the generator. We develop an objective criterion to assess the geenrator performance in a quantitative way. With further development, a generalization of this approach could substantially reduce the needed amount of centrally produced fully simulated events in large particle physics experiments.