National Repository of Dissertations in Serbia
    • English
    • Српски
    • Српски (Serbia)
  • English 
    • English
    • Serbian (Cyrilic)
    • Serbian (Latin)
  • Login
View Item 
  •   NaRDuS home
  • Универзитет у Крагујевцу
  • Економски факултет
  • View Item
  •   NaRDuS home
  • Универзитет у Крагујевцу
  • Економски факултет
  • View Item
JavaScript is disabled for your browser. Some features of this site may not work without it.

Multivarijaciono statističko modeliranje u funkciji merenja stepena ekonomske razvijenosti teritorijalnih jedinica

Thumbnail
2019
Disertacija.pdf (9.250Mb)
Izvestaj_Milan_Stamenkovic_Ekonomski.pdf (736.0Kb)
Author
Stamenković, Milan
Mentor
Savić, Mirko
Committee members
Veselinović, Petar
Metadata
Show full item record
Abstract
U doktorskoj disertaciji su razmatrana suštinska teorijska određenja odabranih multivarijacionih statističkih metoda međuzavisnosti i zavisnosti i sagledani njihovi aplikativni potencijali za modeliranje kompleksnih, multidimenzionih ekonomskih fenomena od interesa, čime je istovremeno opredeljen predmet istraživanja. Afirmaciju primene multivarijacionih statističkih metoda u domenu ekonomskih istraživanja odražava osnovni cilj disertacije, koji podrazumeva kreiranje inovativnog konceptualno-metodološkog okvira, zasnovanog na statistički validnoj implementaciji kako pojedinačnih metoda multivarijacione analize, tako i njihove kombinacije na određenom broju relevantnih pokazatelja, u funkciji merenja dostignutog stepena ekonomske razvijenosti i, shodno tome, klasifikacije teritorijalnih jedinica lokalne samouprave u Republici Srbiji. U tom smislu, u okviru svake metode detaljno su analizirani ciljevi, tipovi i postupak sprovođenja, uz jasno razgraničenje istraživačkih okolnost...i pod kojima se njihova primena smatra prikladnom i statistički opravdanom. Na temeljima važnosti adekvatne pripreme podataka za sprovođenje bilo koje analize podataka u kontekstu obezbeđivanja naučne zasnovanosti dobijenih rezultata i izvedenih zaključaka, posebna pažnja je posvećena pretprocesiranju multivarijacionih opservacija i elaboriranju značaja ispunjenosti statističkih pretpostavki iz perspektive validne primene konkretne metode. Rasvetljavanje kompleksnog i značajnog pitanja validacije kvaliteta rezultata multivarijacionog modeliranja i, s tim u vezi, pronalaženja „statističkih“ argumenta za izbor optimalnog rešenja, izvršeno je analizom brojnih kriterijuma i metoda za evaluaciju rezultata. U empirijskom delu disertacije predstavljena su dva originalna konceptualnometodološka okvira analize multivarijacionih podataka, i to: prvi, zasnovan na integrisanoj primeni faktorske analize, analize grupisanja i multivarijacione analize varijanse u funkciji razvoja multivarijacionog modela (forma kompozitnog pokazatelja) za merenje stepena ekonomske razvijenosti i klasifikaciju jedinica lokalne samouprave u Republici Srbiji, i, drugi, zasnovan na primeni diskriminacione analize u funkciji razvoja klasifikacionog modela za razvrstavanje analiziranih teritorijalnih jedinica u jednu od, prema vrednostima prethodno utvrđenog kompozitnog pokazatelja stepena ekonomske razvijenosti, empirijski identifikovanih grupa. Rezultati istraživanja ukazuju na veliki potencijal kombinovane implementacije multivarijacionih statističkih metoda u koncipiranju inovativnih metodoloških rešenja za analizu i razumevanje ekonomskih fenomena.

In this doctoral dissertation, the essential theoretical determinations of selected multivariate statistical methods of interdependence and dependence were examined, as well as their application potentials for modeling complex, multidimensional economic phenomena of interest were considered, which simultaneously defined the research subject. The affirmation of the application of multivariate statistical methods in the domain of economic research reflects the primary objective of the dissertation, which implies the development of an innovative conceptual-methodological framework, based on statistically valid implementation of individual methods of multivariate analysis, as well as their combinations on a number of relevant indicators in function of measuring the achieved degree of economic development and, accordingly, the classification of local self-government territorial units in the Republic of Serbia. In this sense, within each method, the objectives, types and implementatio...n procedures have been thoroughly analyzed, with a clear distinction of research circumstances under which their application is considered appropriate and statistically justified. On the basis of the importance of adequate data preparation for implementation of any data analysis, in terms of ensuring the scientific basis of the obtained results and conclusions drawn, special attention has been devoted to preprocessing of multivariate observations and elaboration of importance of fulfilling statistical assumptions from the perspective of valid application of particular method. The clarification of complex and significant question of validating the quality of multivariate modeling results and, in this regard, finding “statistical” arguments for choosing the optimal solution, was done by analyzing a number of different criteria and methods for evaluation of results. Within the empirical part of the dissertation, the following two original conceptualmethodological frameworks of multivariate data analysis were presented: first, based on the integrated implementation of factor analysis, cluster analysis and multivariate analysis of variance in function of developing a specific multivariate model (in form of a composite indicator) for measuring the degree of economic development and classification of local selfgovernment units in the Republic of Serbia, and, second, based on the implementation of discriminant analysis in function of developing a classification model for allocation of analyzed territorial units into one of the empirically identified groups, according to the values of the previously proposed composite indicator of degree of economic development. The results of the conducted research indicate the great potential of the combined implementation of multivariate statistical methods in conceiving innovative methodological solutions for the analysis and understanding of economic phenomena.

Faculty:
University of Kragujevac, Faculty of Economics
Date:
18-09-2019
Keywords:
multivarijaciona statistička analiza / multivariate statistical analysis / faktorska analiza / analiza grupisanja / multivarijaciona analiza varijanse / diskriminaciona analiza / kompozitni pokazatelj / stepen ekonomske razvijenosti / regionalni dispariteti / factor analysis / cluster analysis / multivariate analysis of variance / discriminant analysis / composite indicator / degree of economic development / regional disparities
[ Google Scholar ]
URI
http://eteze.kg.ac.rs/application/showtheses?thesesId=7035
http://nardus.mpn.gov.rs/handle/123456789/11714
https://fedorakg.kg.ac.rs/fedora/get/o:1174/bdef:Content/download

Related items

Showing items related by title, author, creator and subject.

  • Razvoj i primena hemometrijskih metoda za klasifikaciju i procenu kvaliteta vode / The development and application of chemometric methods for the classification and assessment of water quality 

    Živojinović, Dragana Z. (University of Belgrade, Faculty of Technology and Metallurgy, 24-09-2013)
  • Hemijski i mikrobiološki profil voda iz artešikih bunara na teritoriji Semberije: analiza faktora koji utiču na mobilnost hemijskih zagađivača 

    Laketić, Tamara J. (University of Niš, Faculty of Science and Mathematics, 06-07-2017)
  • Eksperimentalno određivanje i termodinamičko modelovanje ravnotežnih dijagrama stanja trojnih sistema Ge-Sb-Ag i Ge-Sb-In / Experimental determination and thermodynamic modeling of the phase diagrams of Ge-Sb-Ag and Ge-Sb-In ternary systems 

    Premović, Milena M. (University of Belgrade, Faculty of Technology and Metallurgy, 27-03-2015)

DSpace software copyright © 2002-2015  DuraSpace
About NaRDus | Contact us

OpenAIRERCUBRODOSTEMPUS
 

 

Browse

All of DSpaceUniversities & FacultiesAuthorsMentorCommittee membersSubjectsThis CollectionAuthorsMentorCommittee membersSubjects

DSpace software copyright © 2002-2015  DuraSpace
About NaRDus | Contact us

OpenAIRERCUBRODOSTEMPUS