prikaz prve stranice dokumenta Improving Higgs boson categorisation with Boosted Decision Trees
  Download
PDF 2.33 MB
master's thesis
Improving Higgs boson categorisation with Boosted Decision Trees
2021. urn:nbn:hr:166:546234
Petković, Andro
University of Split
Faculty of Science
Department of Physics

Cite this document

Petković, A. (2021). Improving Higgs boson categorisation with Boosted Decision Trees (Master's thesis). Split: University of Split, Faculty of Science. Retrieved from https://urn.nsk.hr/urn:nbn:hr:166:546234

Petković, Andro. "Improving Higgs boson categorisation with Boosted Decision Trees." Master's thesis, University of Split, Faculty of Science, 2021. https://urn.nsk.hr/urn:nbn:hr:166:546234

Petković, Andro. "Improving Higgs boson categorisation with Boosted Decision Trees." Master's thesis, University of Split, Faculty of Science, 2021. https://urn.nsk.hr/urn:nbn:hr:166:546234

Petković, A. (2021). 'Improving Higgs boson categorisation with Boosted Decision Trees', Master's thesis, University of Split, Faculty of Science, accessed 09 July 2024, https://urn.nsk.hr/urn:nbn:hr:166:546234

Petković A. Improving Higgs boson categorisation with Boosted Decision Trees [Master's thesis]. Split: University of Split, Faculty of Science; 2021 [cited 2024 July 09] Available at: https://urn.nsk.hr/urn:nbn:hr:166:546234

A. Petković, "Improving Higgs boson categorisation with Boosted Decision Trees", Master's thesis, University of Split, Faculty of Science, Split, 2021. Available at: https://urn.nsk.hr/urn:nbn:hr:166:546234

Please login to the repository to save this object to your list.
Metadata
TitleImproving Higgs boson categorisation with Boosted Decision Trees
Title (english)Unaprjeđenje kategorizacije Higgsovog bozona koristeći ubrzana stabla odluke
AuthorAndro Petković
MentorToni Šćulac (mentor)
Committee memberToni Šćulac (predsjednik povjerenstva)
Committee memberMarko Kovač (član povjerenstva)
Committee memberPetar Stipanović (član povjerenstva)
GranterUniversity of Split
Faculty of Science
(Department of Physics)
Split
Defense date and country2021-09-30, Croatia
Scientific / art field,
discipline and subdiscipline		NATURAL SCIENCES
Physics
Abstract
The Higgs boson was discovered in 2012 by CMS and ATLAS Collaborations at the Large Hadron Collider in CERN. Since then, properties of Higgs boson are being measured and within the uncertainity intervals, the latest analysis results are in agreement with predictions of the Standard model (SM). However, more precise measurements have to be performed to establish whether this particle has all the properties of the SM Higgs. In this thesis, two categorizations of Higgs boson production mechanisms are presented. The first categorization is already implemented in CMS analysis. The second categorization, based on a machine learning algorithm - boosted decision tree (BDT), is developed as a part of this project. The improvement in the categorization efficiency has been achieved and the BDT categorization outperformed the existing classification by 6-18%, depending on a specific production mechanism. Using a combination of boosted decision tree and existing classification, signal strength fit was performed on Asimov dataset. Expected value of the signal strength modifier has been determined as well as the corresponding standard deviation.
Abstract (croatian)
Higgsov bozon otrkrile su CMS i ATLAS kolaboracije 2012. godine na CERN-u. Od tada su napravljena razna mjerenja njegovih svojstava i sva su se pokazala konzistentnima s predviđanjima Standardnog Modela unutar intervala pogreške. Međutim, potrebno je napraviti preciznija mjerenja kako bi se odredilo je li čestica uistinu Higgsov bozon iz Standardnog Modela. Svako odstupanje od očekivanih vrijednosti otvorilo bi prozor u novu fiziku. Higgsov bozon može se proizvesti na više načina u velikom hadronskom sudaraču. U ovom diplomskom radu predstavljene su dvije kategorizacije produkcijskih mehanizama Higgsovog bozona. Prva je već postojeća kategorizacija koja je otprije implementirana u CMS analizi. Druga kategorizacija, bazirana na algoritmu strojnog učenja - ubrzanom stablu odluke, razvijena je u sklopu ovog diplomskog rada. Poboljšanje u klasifikaciji je ostvareno te je ubrzano stablo odluke povećalo čistoću kategorija u rasponu 6-18%, ovisno o promatranom produkcijskom mehanizmu, u odnosu na postojeću kategorizaciju. Koristeći kombinaciju klasifikacije bazirane na ubrzanom stablu odluke i otprije implementirane klasifikacije, određena je očekivana vrijednost modifikatora jačine signala te pripadajuća standardna devijacija na Asimovom skupu podataka.
Keywords
Keywords (croatian)
Languageenglish
URN:NBNurn:nbn:hr:166:546234
Study programmeTitle: Physics; specializations in: major in teaching, major in Astrophysics, major in Biophysics, major in Enviromental physics, major in Computational physics, major in Astrophysics and physics of elemental particles
Course: major in Astrophysics and physics of elemental particles
Study programme type: university
Study level: graduate
Academic / professional title: magistar/magistra fizike (magistar/magistra fizike)
Type of resourceText
File originBorn digital
Access conditionsOpen access
Terms of useLicence Attribution-NoDerivatives 4.0 International (CC BY-ND 4.0) icon
Created on2021-10-22 11:31:36