Loss function to optimise signal significance in particle physics

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Bardhan, Jai, Neeraj, Cyrin, Mitra, Subhadip, Mandal, Tanumoy
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910743362797568
author Bardhan, Jai
Neeraj, Cyrin
Mitra, Subhadip
Mandal, Tanumoy
author_facet Bardhan, Jai
Neeraj, Cyrin
Mitra, Subhadip
Mandal, Tanumoy
contents We construct a surrogate loss to directly optimise the significance metric used in particle physics. We evaluate our loss function for a simple event classification task using a linear model and show that it produces decision boundaries that change according to the cross sections of the processes involved. We find that the models trained with the new loss have higher signal efficiency for similar values of estimated signal significance compared to ones trained with a cross-entropy loss, showing promise to improve sensitivity of particle physics searches at colliders.
format Preprint
id arxiv_https___arxiv_org_abs_2412_09500
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Loss function to optimise signal significance in particle physics
Bardhan, Jai
Neeraj, Cyrin
Mitra, Subhadip
Mandal, Tanumoy
High Energy Physics - Phenomenology
Machine Learning
High Energy Physics - Experiment
We construct a surrogate loss to directly optimise the significance metric used in particle physics. We evaluate our loss function for a simple event classification task using a linear model and show that it produces decision boundaries that change according to the cross sections of the processes involved. We find that the models trained with the new loss have higher signal efficiency for similar values of estimated signal significance compared to ones trained with a cross-entropy loss, showing promise to improve sensitivity of particle physics searches at colliders.
title Loss function to optimise signal significance in particle physics
topic High Energy Physics - Phenomenology
Machine Learning
High Energy Physics - Experiment
url https://arxiv.org/abs/2412.09500