Skip to main content
U.S. Department of Energy
Office of Scientific and Technical Information

LeptonJet Categorization with Decision Tree

Conference ·
OSTI ID:1565937

With much evidence for dark matter arising from astronomical observations, particle physics has also begun to look for dark matter candidates. Here, a boosted decision tree machine learning algorithm (BDT) is used to distinguish signal from background in a search for self-interacting dark matter at CMS with two different final states of signal. The signal final states are two displaced collimated pairs of leptons (leptonJets). In this project, possible improvements to the BDT are explored. Various methods of signal separation are studied, as well as various accuracy metrics to study and compare the different models. Through the separated models and improvements in training, a slight increase in accuracy score was obtained (0.01% on electrons, 0.46% on muons). Additionally, the separated models were able to recognize a significantly increased percentage of signal, particularly against the electron signal. Separated models are being considered as part of the larger an alysis.

Research Organization:
Fermi National Accelerator Laboratory (FNAL), Batavia, IL (United States)
Sponsoring Organization:
USDOE Office of Science (SC), High Energy Physics (HEP) (SC-25)
Contributing Organization:
CMS
DOE Contract Number:
AC02-07CH11359
OSTI ID:
1565937
Report Number(s):
FERMILAB-POSTER-19-118-CMS; oai:inspirehep.net:1754619
Country of Publication:
United States
Language:
English

Similar Records

Fast inference of Boosted Decision Trees in FPGAs for particle physics
Journal Article · Fri May 29 00:00:00 EDT 2020 · Journal of Instrumentation · OSTI ID:1647080

Using Boosted Decision Trees to Separate Signal and Background in B to XsGamma Decays
Technical Report · Wed Sep 27 00:00:00 EDT 2006 · OSTI ID:892609

Search for resonant production of strongly coupled dark matter in proton-proton collisions at 13 TeV
Journal Article · Wed Jun 29 00:00:00 EDT 2022 · Journal of High Energy Physics (Online) · OSTI ID:1838468