Transforming Simulation to Data Without Pairing

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Gendreau-Distler, Eli, Pottier, Luc Le, Wang, Haichen
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910912986742784
author Gendreau-Distler, Eli
Pottier, Luc Le
Wang, Haichen
author_facet Gendreau-Distler, Eli
Pottier, Luc Le
Wang, Haichen
contents We explore a generative machine learning-based approach for estimating multi-dimensional probability density functions (PDFs) in a target sample using a statistically independent but related control sample - a common challenge in particle physics data analysis. The generative model must accurately reproduce individual observable distributions while preserving the correlations between them, based on the input multidimensional distribution from the control sample. Here we present a conditional normalizing flow model (CNF) based on a chain of bijectors which learns to transform unpaired simulation events to data events. We assess the performance of the CNF model in the context of LHC Higgs to diphoton analysis, where we use the CNF model to convert a Monte Carlo diphoton sample to one that models data. We show that the CNF model can accurately model complex data distributions and correlations. We also leverage the recently popularized Modified Differential Multiplier Method (MDMM) to improve the convergence of our model and assign physical meaning to usually arbitrary loss-function parameters.
format Preprint
id arxiv_https___arxiv_org_abs_2504_12343
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Transforming Simulation to Data Without Pairing
Gendreau-Distler, Eli
Pottier, Luc Le
Wang, Haichen
Data Analysis, Statistics and Probability
High Energy Physics - Experiment
High Energy Physics - Phenomenology
We explore a generative machine learning-based approach for estimating multi-dimensional probability density functions (PDFs) in a target sample using a statistically independent but related control sample - a common challenge in particle physics data analysis. The generative model must accurately reproduce individual observable distributions while preserving the correlations between them, based on the input multidimensional distribution from the control sample. Here we present a conditional normalizing flow model (CNF) based on a chain of bijectors which learns to transform unpaired simulation events to data events. We assess the performance of the CNF model in the context of LHC Higgs to diphoton analysis, where we use the CNF model to convert a Monte Carlo diphoton sample to one that models data. We show that the CNF model can accurately model complex data distributions and correlations. We also leverage the recently popularized Modified Differential Multiplier Method (MDMM) to improve the convergence of our model and assign physical meaning to usually arbitrary loss-function parameters.
title Transforming Simulation to Data Without Pairing
topic Data Analysis, Statistics and Probability
High Energy Physics - Experiment
High Energy Physics - Phenomenology
url https://arxiv.org/abs/2504.12343