Punzi-loss: a non-differentiable metric approximation for sensitivity optimisation in the search for new particles
The result's identifiers
Result code in IS VaVaI
<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216208%3A11320%2F22%3A10455205" target="_blank" >RIV/00216208:11320/22:10455205 - isvavai.cz</a>
Result on the web
<a href="https://verso.is.cuni.cz/pub/verso.fpl?fname=obd_publikace_handle&handle=nD49R5Bfnc" target="_blank" >https://verso.is.cuni.cz/pub/verso.fpl?fname=obd_publikace_handle&handle=nD49R5Bfnc</a>
DOI - Digital Object Identifier
<a href="http://dx.doi.org/10.1140/epjc/s10052-022-10070-0" target="_blank" >10.1140/epjc/s10052-022-10070-0</a>
Alternative languages
Result language
angličtina
Original language name
Punzi-loss: a non-differentiable metric approximation for sensitivity optimisation in the search for new particles
Original language description
We present the novel implementation of a non-differentiable metric approximation and a corresponding loss-scheduling aimed at the search for new particles of unknown mass in high energy physics experiments. We call the loss-scheduling, based on the minimisation of a figure-of-merit related function typical of particle physics, a Punzi-loss function, and the neural network that utilises this loss function a Punzi-net. We show that the Punzi-net outperforms standard multivariate analysis techniques and generalises well to mass hypotheses for which it was not trained. This is achieved by training a single classifier that provides a coherent and optimal classification of all signal hypotheses over the whole search space. Our result constitutes a complementary approach to fully differentiable analyses in particle physics. We implemented this work using PyTorch and provide users full access to a public repository containing all the codes and a training example.
Czech name
—
Czech description
—
Classification
Type
J<sub>imp</sub> - Article in a specialist periodical, which is included in the Web of Science database
CEP classification
—
OECD FORD branch
10300 - Physical sciences
Result continuities
Project
—
Continuities
I - Institucionalni podpora na dlouhodoby koncepcni rozvoj vyzkumne organizace
Others
Publication year
2022
Confidentiality
S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů
Data specific for result type
Name of the periodical
European Physical Journal C
ISSN
1434-6044
e-ISSN
1434-6052
Volume of the periodical
82
Issue of the periodical within the volume
2
Country of publishing house
DE - GERMANY
Number of pages
8
Pages from-to
121
UT code for WoS article
000752936500003
EID of the result in the Scopus database
2-s2.0-85124749815