prikaz prve stranice dokumenta Prepoznavanje malicioznih napada pomoću strojnog učenja
  Download
PDF 1.26 MB
undergraduate thesis
Prepoznavanje malicioznih napada pomoću strojnog učenja
2024. urn:nbn:hr:200:721961
Carević, Antonio
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

Cite this document

Carević, A. (2024). Prepoznavanje malicioznih napada pomoću strojnog učenja (Undergraduate thesis). Osijek: Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek. Retrieved from https://urn.nsk.hr/urn:nbn:hr:200:721961

Carević, Antonio. "Prepoznavanje malicioznih napada pomoću strojnog učenja." Undergraduate thesis, Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek, 2024. https://urn.nsk.hr/urn:nbn:hr:200:721961

Carević, Antonio. "Prepoznavanje malicioznih napada pomoću strojnog učenja." Undergraduate thesis, Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek, 2024. https://urn.nsk.hr/urn:nbn:hr:200:721961

Carević, A. (2024). 'Prepoznavanje malicioznih napada pomoću strojnog učenja', Undergraduate thesis, Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek, accessed 21 February 2025, https://urn.nsk.hr/urn:nbn:hr:200:721961

Carević A. Prepoznavanje malicioznih napada pomoću strojnog učenja [Undergraduate thesis]. Osijek: Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek; 2024 [cited 2025 February 21] Available at: https://urn.nsk.hr/urn:nbn:hr:200:721961

A. Carević, "Prepoznavanje malicioznih napada pomoću strojnog učenja", Undergraduate thesis, Josip Juraj Strossmayer University of Osijek, Faculty of Electrical Engineering, Computer Science and Information Technology Osijek, Osijek, 2024. Available at: https://urn.nsk.hr/urn:nbn:hr:200:721961

Please login to the repository to save this object to your list.
Metadata
TitlePrepoznavanje malicioznih napada pomoću strojnog učenja
Title (english)Malicious attack detection using machine learning
AuthorAntonio Carević
MentorDražen Bajer (mentor)
MentorMario Dudjak (sumentor)
Committee memberRobert Cupec (predsjednik povjerenstva)
GranterJosip 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)
Osijek
Defense date and country2024-09-26, Croatia
Scientific / art field,
discipline and subdiscipline		TECHNICAL SCIENCES
Computing
Artificial Intelligence
Abstract
Cilj ovog rada bio je odrediti i primijeniti tok učenja odabranih algoritama klasifikacije iz skupova podataka koji opisuju različite vrste malicioznih napada i definirati procedure unutar tog toka. Opisani su maliciozni napadi koji predstavljaju prijetnju kao što su trojanski konj, mrežna krađa identiteta, virus i distribuirani napadi uskraćivanjem usluga te se pruža uvid u tradicionalne načine obrane od ovih napada. Osim toga, objašnjen je pojam strojnog učenja i zadatak klasifikacije kao posebne vrste nadziranog strojnog učenja. Pregledom literature, prikazani su problemi u kojima je korišteno strojno učenje za izgradnju sustava za obranu od malicioznih napada. Analizirani su klasifikacijski algoritmi korišteni u radu te su predstavljene tehnike za odabir značajki i rad sa nebalansiranim skupovima podataka. U posljednjem dijelu rada predstavljene su informacije o korištenim skupovima podataka i objašnjene su postavke i metodologije u procesu učenja. Procijenjena je efikasnost metoda za odabir značajki i tehnika uzorkovanja te je analiziran njihov doprinos poboljšanju modela. Prilikom klasifikacije malicioznih napada najbolje rezultate ostvario je algoritam stablo odluke . Nasuprot njemu, najlošije se pokazao algoritam Naive Bayes. Upotrebom tehnika za odabir značajki modeli ostvaruju jednako dobre ili bolje performanse smanjenjem broja značajki. SFS omotač je jedna od tehnika koja se ističe zbog doprinosa u poboljšanju modela. Smanjenjem stupnja neuravnoteženosti skupova podataka performanse rastu upotrebom tehnike nasumičnog preuzorkovanja.
Abstract (english)
The goal of this work was to determine and apply the learning flow of selected classification algorithms from data sets describing different types of malicious attacks and to define the procedures within that flow. Malicious attacks such as Trojan horses, phishing, viruses, and distributed denial-ofservice attacks are described, and traditional defenses against these attacks are provided. In addition, the concept of machine learning and the task of classification as a special type of supervised machine learning are explained. By reviewing the literature, problems in which machine learning was used to build a system to defend against malicious attacks were presented. The classification algorithms used in the work were analyzed and the techniques for features selection and handeling imbalanced data sets were presented. In the last part, information on the used data sets is presented and the settings and methodologies in the learning process are explained. The effectiveness of feature selection methods and sampling techniques was evaluated and their contribution to model improvement was analyzed. When classifying malicious attacks, the decision tree algorithm achieved the best results. In contrast, the Naive Bayes algorithm performed the worst. By using feature selection techniques, models achieve equal or better performance by reducing the number of features. The SFS wrapper is one of the techniques that stands out for its contribution to model improvement. By reducing the imbalance ratio of data sets, performance increases using the random oversampling technique.
Keywords
Keywords (english)
Languagecroatian
URN:NBNurn:nbn:hr:200:721961
Study programmeTitle: Undergraduate University Study Programme in Computer Engineering
Study programme type: university
Study level: undergraduate
Academic / professional title: sveučilišni/a prvostupnik/prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka računarstva (sveučilišni/a prvostupnik/prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka računarstva)
Type of resourceText
File originBorn digital
Access conditionsOpen access
Terms of useLicence In copyright icon
Created on2024-10-14 08:09:04