prikaz prve stranice dokumenta Problem multikolinearnosti u višestrukoj regresiji: detekcija i moguća rješenja
  Download
PDF 446.84 KB
master's thesis
Problem multikolinearnosti u višestrukoj regresiji: detekcija i moguća rješenja
2023. urn:nbn:hr:166:176287
Nakić, Jelena
University of Split
Faculty of Science
Department of Mathematics

Cite this document

Nakić, J. (2023). Problem multikolinearnosti u višestrukoj regresiji: detekcija i moguća rješenja (Master's thesis). Split: University of Split, Faculty of Science. Retrieved from https://urn.nsk.hr/urn:nbn:hr:166:176287

Nakić, Jelena. "Problem multikolinearnosti u višestrukoj regresiji: detekcija i moguća rješenja." Master's thesis, University of Split, Faculty of Science, 2023. https://urn.nsk.hr/urn:nbn:hr:166:176287

Nakić, Jelena. "Problem multikolinearnosti u višestrukoj regresiji: detekcija i moguća rješenja." Master's thesis, University of Split, Faculty of Science, 2023. https://urn.nsk.hr/urn:nbn:hr:166:176287

Nakić, J. (2023). 'Problem multikolinearnosti u višestrukoj regresiji: detekcija i moguća rješenja', Master's thesis, University of Split, Faculty of Science, accessed 15 April 2024, https://urn.nsk.hr/urn:nbn:hr:166:176287

Nakić J. Problem multikolinearnosti u višestrukoj regresiji: detekcija i moguća rješenja [Master's thesis]. Split: University of Split, Faculty of Science; 2023 [cited 2024 April 15] Available at: https://urn.nsk.hr/urn:nbn:hr:166:176287

J. Nakić, "Problem multikolinearnosti u višestrukoj regresiji: detekcija i moguća rješenja", Master's thesis, University of Split, Faculty of Science, Split, 2023. Available at: https://urn.nsk.hr/urn:nbn:hr:166:176287

Please login to the repository to save this object to your list.
Metadata
TitleProblem multikolinearnosti u višestrukoj regresiji: detekcija i moguća rješenja
Title (english)Multicollinearity in multiple regression: detection methods and possible solutions
AuthorJelena Nakić
MentorJurica Perić (mentor)
MentorAna Perišić (pomoć pri izradi, neposredni voditelj)
Committee memberJurica Perić (predsjednik povjerenstva)
Committee memberAna Perišić (član povjerenstva)
Committee memberMarcela Mandarić (član povjerenstva)
GranterUniversity of Split
Faculty of Science
(Department of Mathematics)
Split
Defense date and country2023-12-22, Croatia
Scientific / art field,
discipline and subdiscipline		NATURAL SCIENCES
Mathematics
Abstract
Linearna regresija je metoda modeliranja veze između varijable odgovora i jedne ili više eksplanatornih varijabli. Jedan od problema koji može nastupiti je slučaj kada su barem dvije eksplanatorne varijable linearno zavisne ili približno linearno zavisne. Kažemo da tada imamo problem multikolinearnosti. U ovom radu navodimo metode za detekciju i rješavanje ovoga problema, a zatim prikazujemo kako ih koristiti na stvarnom skupu podataka. Gradimo različite regresijske modele na kojima provjeravamo prisutnost problema multikolinearnosti te ukoliko je prisutan rješavamo ga izostavljanjem varijabli i analizom glavnih komponenti. Pomoću obje metode možemo doći do modela bez problema multikolinearnosti, ali jasnija je interpretacija modela dobivenog izostavljanjem varijabli.
Abstract (english)
Linear regression is a method of modeling the relationship between a response variable and one or more explanatory variables. One of the problems that can occur is the case when at least two explanatory variables are linearly dependent or approximately linearly dependent. In this case, we say that the problem of multicollinearity is present. In this paper, we present several methods for detecting and solving this problem, and then present the application on a real data set. We build different regression models where we check the presence of multicollinearity. We tackle the presence of multicollinearity by omitting variables or applying principal component analysis. In each case it is possible to build a model having no serious problem of multicollinearity. However, the model built on the basis of principal components offers limited interpretation.
Keywords
Keywords (english)
Languagecroatian
URN:NBNurn:nbn:hr:166:176287
Study programmeTitle: Mathematics; specializations in: major in pure mathematics, major in computing, major in teaching, major in statistics and computing
Course: major in statistics and computing
Study programme type: university
Study level: graduate
Academic / professional title: magistar/magistra matematike (magistar/magistra matematike)
Type of resourceText
File originBorn digital
Access conditionsOpen access
Terms of useLicence In copyright icon
Created on2024-01-15 13:22:20