A comparison of methodologies with minimal hyper-parameter tuning for anomaly detection on transactional data

Simple item page
dc.contributor.advisorVisagie, Stephan Esterhuyseen_ZA
dc.contributor.authorAlderslade, James Williamen_ZA
dc.contributor.otherStellenbosch University. Faculty of Economic and Management Sciences. Dept. of Logistics.en_ZA
dc.date.accessioned2023-03-03T14:08:53Z
dc.date.accessioned2023-05-18T07:05:54Z
dc.date.available2023-03-03T14:08:53Z
dc.date.available2023-05-18T07:05:54Z
dc.date.issued2023-03
dc.descriptionThesis (MCom)--Stellenbosch University, 2023.en_ZA
dc.description.abstractENGLISH SUMMARY: In the information technology era where data has become as valuable as gold, being able to process and use it effectively is vital. Many large industries ranging from healthcare and insurance to banking utilise some form of transactional data in order to track and process many of the services and products which they offer to their clients. These processes are moving too fast for any human to keep up, due to the ever increasing rates of service and convenience which are core drivers in this technological era. Under these circumstances, anomaly detection is required. Fraud detection is the process whereby these transactions are identified and so are the items which could be related to any fraudulent, wasteful or abusive behaviour. A variety of anomaly detection methodologies are investigated to illustrate and compare their ability to quickly and accurately detect anomalous transactions without being given a set of rules. The methods used range from traditional statistical methods, through to machine learning and deep learning. Isolation forest performs best on the evaluation criteria used in this study.en_ZA
dc.description.abstractAFRIKAANSE OPSOMMING: In die inligtingstegnologie-era waar data so waardevol soos goud geword het, is dit noodsaaklik om dit doeltreffend te verwerk en te gebruik. Baie groot nywerhede wat wissel van gesondheidsorg en versekering tot bankwese gebruik een of ander vorm van transaksiedata om baie van die dienste en produkte wat hulle aan hul kliente bied, op te spoor en te verwerk. Hierdie prosesse beweeg te vinnig vir enige mens om by te hou, as gevolg van die steeds toenemende tempo van diens en gerieflewering wat die kerndrywers in hierdie tegnologiese era is. Omstandighede vereis anomalie opsporing. Bedrogopsporing is die proses waardeur hierdie transaksies uitgelig word en so ook die items wat verband kan hou met enige bedrieglike, verkwistende of beledigende gedrag. ’n Verskeidenheid anomalie-opsporingsmetodologiee word ondersoek om hul vermoe om vinnig en akkuraat afwykende transaksies op te tel, te illustreer en te vergelyk. Die metodes wat gebruik word wissel van tradisionele statistiese metodes tot masjienleer en diep leer. Isolasie woude presteer die beste met die evalueringskriteria kriteria wat in hierdie studie gebruik word.af_ZA
dc.description.versionMasters
dc.format.extentxii, 84 pages : illustrations
dc.identifier.urihttp://hdl.handle.net/10019.1/127133
dc.language.isoen_ZAen_ZA
dc.publisherStellenbosch : Stellenbosch University
dc.rights.holderStellenbosch University
dc.subject.lcshAnomaly detection (Computer security)en_ZA
dc.subject.lcshMachine learning -- Security measuresen_ZA
dc.subject.lcshComputer networks -- Reliabilityen_ZA
dc.subject.nameUCTD
dc.titleA comparison of methodologies with minimal hyper-parameter tuning for anomaly detection on transactional dataen_ZA
dc.typeThesisen_ZA
Files
Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
alderslade_comparison_2023.pdf
Size:
2.27 MB
Format:
Adobe Portable Document Format
Description:
Download
Collections
Masters Degrees (Logistics)