Korištenje novih reprezentacija podataka u svrhu klasfikacije događaja pomoću strojnog učenja

master's thesis

• Title (english): Usage of new data representations for event classification with machine learning methods
• Author: Nino Kovačić
• Mentor: Nikola Poljak (mentor)
• Committee member: Nikola Poljak (predsjednik povjerenstva)
• Committee member: Davor Horvatić (član povjerenstva)
• Committee member: Krešimir Kumerički (član povjerenstva)
• Committee member: Darko Androić (član povjerenstva)
• Committee member: Andrej Novak (član povjerenstva)
• Granter: University of Zagreb; Faculty of Science; (Department of Physics); Zagreb
• Defense date and country: 2022-11-24, Croatia
• Scientific / art field, discipline and subdiscipline: NATURAL SCIENCES; Physics

Abstract

Kako bi se nosili sa stalno povećavajućim količinama podataka generiranih u eksperimentima potrebno je primijeniti naprednije tehnike obrade podataka. Najbolji primjer kojih je strojno učenje, ili specifičnije, neuralne mreže. Koristeći generirane podatke o događajima s mlazovima razvijamo binarni klasifikator s ciljem razlikovanja potječe li mlaz od gluona ili kvarka. U radu proučit ćemo nekoliko reprezentacija podataka te različite arhitekture mreža i usporediti klasifikatore.

Abstract (english)

In order to deal with the constantly increasing amount of data generated by experiments more advanced data processing techniques are required. The best example of these is machine learning, and more specifically, neural networks. Using a generated ”black box” set of events describing jets, we develop a binary classifier to determine whether the origin of a particular jet is a quark or a gluon. In doing so, we’ll examine several representations of events as well as different network architectures, and compare the classifiers.

• Keywords: klasifikacija događaja pomoću strojnog učenja
• Language: croatian
• URN:NBN: urn:nbn:hr:217:361469
• Study programme: Title: Physics; specializations in: Research; Course: Research; Study programme type: university; Study level: integrated undergraduate and graduate; Academic / professional title: magistar/magistra fizike (magistar/magistra fizike)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Open access
• Created on: 2022-12-19 13:21:36