Detekcija anomalija u grafovima

master's thesis

• Title (english): Anomaly detection in graphs
• Author: Dominik Galić
• Mentor: Boris Vrdoljak (mentor)
• Committee member: Boris Vrdoljak (predsjednik povjerenstva)
• Committee member: Mario Brčić (član povjerenstva)
• Committee member: Mirta Baranović (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2021-07-06, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

U ovom diplomskom radu prikazane su vrste grafova s obzirom na vremensku promjenjivost topologije, najčešće vrste anomalija te metode za otkrivanje anomalija u grafovima. Prvi korak je istraživanje postojećih implementacija nad popularnim skupovima podataka koji su javno dostupni. Nakon toga izabrana su dva pristupa za detekciju anomalija u anonimiziranim podacima telekomunikacijske kompanije. Prvi pristup detektirao je anomalije koristeći metodu dubokog strojnog učenja autoencodere, dok se drugi pristup temelji na algoritmu DNODA koji promatra direktno susjedstvo čvorova u grafu. Metode su implementirane korištenjem programskog jezika Python, a vizualizacije bibliotekama za vizualizaciju jezika R.

Abstract (english)

Within this thesis, I have shown different types of graphs based on their topology with respect to the temporal component, the most common types of graph anomalies, and methods for detecting anomalies in graphs. The first step is to investigate existing implementations on popular datasets. The next step is choosing two methods for graph anomaly detection in the anonymized dataset from a telecommunication company. The first approach is detecting anomalies using the deep learning method – autoencoders, and the second approach is based on detecting anomalies with respect to the direct neighborhood of graph nodes. Algorithms were implemented using Python programming language and visualizations are made with programming libraries for the R language.

• Keywords: graf; anomalija; detekcija; Python; vizualizacija
• Keywords (english): graph; anomaly; detection; Python; visualization
• Language: croatian
• URN:NBN: urn:nbn:hr:168:362565
• 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: 2023-01-13 14:55:42