Algoritmi spektralnog grupiranja zasnovanog na motivima

master's thesis

• Title (english): Motif-Based Spectral Clustering Algorithms
• Author: Antonio Šoštar
• Mentor: Igor Čavrak (mentor)
• Committee member: Igor Čavrak (predsjednik povjerenstva)
• Committee member: Danko Basch (član povjerenstva)
• Committee member: Tomislav Lipić (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2019-07-15, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

Cilj rada je prilagoditi algoritam spektralnog grupiranja zasnovanog na motivima izvršavanju na raspodijeljenom okruženju (na računalnom grozdu) pomoću radnog okvira Apache Spark. Algoritam se sastoji od dvije faze, koje su računski podjednako zahtjevne - računanje matrice sličnosti i spektralno grupiranje vrhova. Ideja algoritma je dobiti grupiranja višeg reda, tako da se prebroji pojavljivanje para vrhova u zadanom motivu u mreži, te tako definirati matricu susjedstva iz čijeg se laplasijana traži optimalni svojstveni vektor. Algoritam je implementiran za radni okvir Apache Spark, te pokazuje visoki stupanj paralelizma i skalabilnosti.

Abstract (english)

The aim of this thesis is to implement a motif based spectral clustering algorithm on a distributed environment (on a computer cluster) using the Apache Spark framework. The algorithm consists of two phases, which are computationally equally demanding - computing the similarity matrix and spectral clustering of nodes. The idea of the algorithm is to get a higher order clustering, defining the neighborhood matrix by counting the pairs of nodes in the given motif in the network, from which Laplacian an optimal vector is computed. The algorithm is implemented in the Apache Spark framework and shows a high degree of parallelism and scalability.

• Keywords: spektralno grupiranje; motivi; Apache Spark
• Keywords (english): spectral clustering; motifs; Apache Spark
• Language: croatian
• URN:NBN: urn:nbn:hr:168:682957
• Study programme: Title: Computing; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra inženjer/inženjerka računarstva (magistar/magistra inženjer/inženjerka računarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2020-01-07 22:29:10