prikaz prve stranice dokumenta Oblikovanje radijalnih mreža za regresiju
  Download
PDF 1.74 MB
master's thesis
Oblikovanje radijalnih mreža za regresiju
2019. urn:nbn:hr:200:822448
Đakovac, Matej
Josip Juraj Strossmayer University of Osijek
Faculty of Electrical Engineering, Computer Science and Information Technology Osijek
Department of Software Engineering
Chair of Programming Languages and Systems

Cite this document

Đakovac, M. (2019). Oblikovanje radijalnih mreža za regresiju (Master's thesis). Retrieved from https://urn.nsk.hr/urn:nbn:hr:200:822448

Đakovac, Matej. "Oblikovanje radijalnih mreža za regresiju." Master's thesis, Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek, 2019. https://urn.nsk.hr/urn:nbn:hr:200:822448

Đakovac, Matej. "Oblikovanje radijalnih mreža za regresiju." Master's thesis, Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek, 2019. https://urn.nsk.hr/urn:nbn:hr:200:822448

Đakovac, M. (2019). 'Oblikovanje radijalnih mreža za regresiju', Master's thesis, Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek, accessed 12 May 2021, https://urn.nsk.hr/urn:nbn:hr:200:822448

Đakovac M. Oblikovanje radijalnih mreža za regresiju [Master's thesis]. Osijek: Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek; 2019 [cited 2021 May 12] Available at: https://urn.nsk.hr/urn:nbn:hr:200:822448

M. Đakovac, "Oblikovanje radijalnih mreža za regresiju", Master's thesis, Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek, Osijek, 2019. Available at: https://urn.nsk.hr/urn:nbn:hr:200:822448

Islandora Bookmark - Please login to use this feature
Metadata
TitleOblikovanje radijalnih mreža za regresiju
Title (english)Designing radial basis function networks for regression
AuthorMatej Đakovac
MentorGoran Martinović (mentor)
MentorDražen Bajer (mentor)
Committee memberGoran Martinović (predsjednik povjerenstva)
Committee memberAlfonzo Baumgartner (član povjerenstva)
Committee memberBruno Zorić (član povjerenstva)
GranterJosip Juraj Strossmayer University of Osijek
Faculty of Electrical Engineering, Computer Science and Information Technology Osijek
(Department of Software Engineering)
(Chair of Programming Languages and Systems)
Osijek
Defense date and country2019-10-10, Croatia
Scientific / art field,
discipline and subdiscipline		TECHNICAL SCIENCES
Computing
Artificial Intelligence
Abstract
U diplomskom radu opisan je problem regresije, te pristup njegovom rješavanju korištenjem radijalnih mreža. Objašnjen je koncept i arhitektura radijalnih mreža i objašnjene su sličnosti i razlike s višeslojnim perceptronom te tijek izgradnje koji se sastoji od određivanja parametara koristeći nadzirane i nenadzirane metode učenja. Navedeni su različiti načini određivanja svakog od parametara. Razvijeno je programsko rješenje koje omogućuje izvršavanje regresijske analize na danom skupu podataka tako da se učita skup podataka, konfigurira radijalna mreža, te koje pruža na uvid rezultate. Izvršena je eksperimentalna analiza korištenjem razvijenog programskog rješenja za regresijsku analizu pet stvarnih skupa podataka čiji su rezultati prikazani i komentirani. Dobiveni rezultati pokazuju važnost korištenja algoritma za grupiranje i skaliranja širina radijalnih funkcija pri određivanju parametara radijalne mreže.
Abstract (english)
Graduation thesis provides a description of regression analysis as well as an approach to regression analysis using radial basis function networks. The general concept and architecture of radial basis function networks are explained, and also the process of modelling such a network which is done by determining its parameters using supervised and unsupervised learning methods. Different approaches to determining the parameters are presented. A CLI application was developed to perform regression analysis, which enables loading of a dataset, network parameter configuration, and which provides results for inspection to the user. Experimental analysis was performed using the application for regression analysis of five real-world datasets, whose results were shown and commented. The results show the importance of clustering algorithm usage and radial basis function width scaling when determining network parameters.
Keywords
Keywords (english)
Languagecroatian
URN:NBNurn:nbn:hr:200:822448
Study programmeTitle: Graduate University Study Programme in Computer Engineering
Study programme type: university
Study level: graduate
Academic / professional title: magistar/magistra inženjer/inženjerka računarstva (magistar/magistra inženjer/inženjerka računarstva)
Type of resourceText
File originBorn digital
Access conditionsOpen access
Terms of useLicence In copyright icon
Created on2019-12-23 07:04:36