Analiza korespodencije

master's thesis

• Author: Vladimir Jerebić
• Mentor: Lavoslav Čaklović (mentor)
• Committee member: Lavoslav Čaklović (predsjednik povjerenstva)
• Committee member: Boris Guljaš (član povjerenstva)
• Committee member: Ante Mimica (član povjerenstva)
• Committee member: Marko Vrdoljak (član povjerenstva)
• Granter: University of Zagreb; Faculty of Science; (Department of Mathematics); Zagreb
• Defense date and country: 2015-07-13, Croatia
• Scientific / art field, discipline and subdiscipline: NATURAL SCIENCES; Mathematics

Abstract

Cilj ovog rada je predstaviti problem vizualizacije višedimenzionalnih podataka i prikazati rješenje problema pomoću analize korespondencije. Analiza korespondencije je statistička metoda za vizualizaciju kategorijskih podataka koju je uveo Jean-Paul Benzécri šezdesetih godina prethodnog stoljeća. U međuvremenu se analiza korespondencije nametnula kao vrlo efikasna metoda za analizu kategorijskih podataka koja se svakim danom sve više koristi u društvenim i prirodnim znanostima. U prvom poglavlju smo se najprije upoznali s biplotom kao osnovnim grafičkim alatom za vizualizaciju višedimenzionalnih podataka. Uz to, objasnili smo postupak konstrukcije različitih vrsta biplota i preko poopćene metode analize glavne komponente stigli do motivacije za analizu korespondencije. U drugom poglavlju smo detaljno opisali osnovnu analizu korespondencije, demonstrirali metodu na nekoliko primjera te uveli dvije modifikacije osnovne metode. Konačno, u trećem poglavlju se nalaze kodovi korišteni pri obradi podataka za ovaj rad.

Abstract (english)

The goal of this thesis was to present the problem of visualisation multidimensional data and provide the solution using correspondence analysis. Correspondence analysis is a statistical technique developed by Jean-Paul Benzécri during the sixties. In the meantime, correspondence analysis has imposed as very effective method for analysis of the categorical data and as such it is used in social and natural sciences more each day. In the first chapter, we introduced the biplot as the basic graphic tool for visualization of multidimensional data. In addition, we described the procedure of biplot construction, explained several types of biplot and using generalised principal component analysis developed motivation for correspondence analysis. In the second chapter, we have described, in detail, simple correspondence analysis and demonstrated method on the several examples. Later in chapter, we introduced two modifications of the simple correspondence analysis. Finally, in the third chapter, we present the R codes used for data analysis in this thesis.

• Keywords: problem vizualizacije; višedimenzionalni podataci; analiza korespondencije; Benzécri; analiza kategorijskih podataka
• Keywords (english): problem of visualisation; multidimensional data; correspondence analysis; Benzécri; analysis of categorical data
• Language: croatian
• URN:NBN: urn:nbn:hr:217:805932
• Study programme: Title: Finance and Business Mathematics; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra matematike (magistar/magistra matematike)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Open access
• Created on: 2019-01-31 13:56:15