master's thesis
Differential evolution algorithm for automatic data clustering

Karlo Kotrba (2015)
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
Metadata
TitleAlgoritam diferencijalne evolucije za automatsko grupiranje podataka
AuthorKarlo Kotrba
Mentor(s)Goran Martinović (thesis advisor)
Dražen Bajer (thesis advisor)
Abstract
Diplomski rad obrađuje uvod u algoritme za grupiranje podataka. Ukratko su obrađeni glavni tipovi algoritama, te načini vrednovanja kvalitete particije. Detaljnije je objašnjena diferencijalna evolucija uz povezivanje sa evolucijskim algoritmima. Prikazani su načini primjene diferencijalne evolucije pri grupiranju podataka u poznat ili nepoznat broj grupa. Također je opisan način pronalaženja odgovarajućeg broja grupa ukoliko on nije unaprijed poznat. Izrađeno je programsko rješenje koje omogućuje usporedbu raznih načina vrednovanja kvalitete grupiranja. Također je provedena eksperimentalna analiza gdje su prikazani i komentirani rezultati usporedbe. Uz ovo analizirana je i efikasnost raznih načina paraleliziranja programskog koda.
Keywordsautomatic data clustering differential evolution data clustering cluster validity index parallel loops.
Parallel title (English)Differential evolution algorithm for automatic data clustering
Committee MembersGoran Martinović (committee chairperson)
Ivica Lukić (committee member)
Dražen Bajer (committee member)
GranterJosip Juraj Strossmayer University of Osijek
Faculty of Electrical Engineering, Computer Science and Information Technology Osijek
Lower level organizational unitsDepartment of Software Engineering
Chair of Programming Languages and Systems
PlaceOsijek
StateCroatia
Scientific field, discipline, subdisciplineTECHNICAL SCIENCES
Computing
Process Computing
Study programme typeuniversity
Study levelgraduate
Study programmeGraduate University Study Programme in Computer Engineering
Academic title abbreviationmag.ing.comp.
Genremaster's thesis
Language Croatian
Defense date2015-10-05
Parallel abstract (English)
The thesis gives an introduction to data clustering algorithms. Main types of data clustering algorithms as well as different clustering validity criteria were described. Differential evolution and its application related to data clustering was described in detail. The applicability of differential evolution for automatic clustering as well as clustering when the number of clusters is known in advance was shown. Automatic clustering was described in depth. Software solution that enables a comparison between different clustering validity criteria was developed. The results of the aforementioned comparison were shown and commented. Efficacy analysis of multiple implementations of parallelism was also shown.
Parallel keywords (Croatian)automatsko grupiranje podataka diferencijalna evolucija grupiranje podataka kriteriji vrednovanja grupiranja paralelne petlje.
Resource typetext
Access conditionOpen access
Terms of usehttp://rightsstatements.org/vocab/InC/1.0/
URN:NBNhttps://urn.nsk.hr/urn:nbn:hr:200:986611
CommitterAnka Ovničević