Krāpniecības signālu identificēšana, izmantojot Beijesa tīklus
Loading...
Date
Authors
Advisor
Journal Title
Journal ISSN
Volume Title
Publisher
Latvijas Universitāte
Language
lav
Abstract
Darbā ir apskatīta Beijesa tīklu pielietojamība krāpniecības signālu identificēšanai, izmantojot klasificēšanas metodi. Aprakstīti Beijesu tīklu strukturēšanas un uzskatu atjaunošanas principi. Izpētīti mainīgo lielumu varbūtību sadalījumu parametru novērtētāji. Apskatīti Beijesa tīklu veidi. Izstrādāti Beijesa tīklu modeļi krāpniecības identificēšanai un iegūtie rezultāti salīdzināti ar citiem mašīnmācīšanās algoritmiem.
The work examines the applicability of Bayesian Networks for fraud identification using a classification approach. It describes the principles of Bayesian network structuring and belief up dating. The study examines parameter estimators for the probability distributions of random variables. Various types of Bayesian Networks have been reviewed. Finally, Bayesian Network models for fraud detection have been developed, and the results obtained have been compared with those of other machine learning algorithms.
The work examines the applicability of Bayesian Networks for fraud identification using a classification approach. It describes the principles of Bayesian network structuring and belief up dating. The study examines parameter estimators for the probability distributions of random variables. Various types of Bayesian Networks have been reviewed. Finally, Bayesian Network models for fraud detection have been developed, and the results obtained have been compared with those of other machine learning algorithms.