Abstract:
Fraudsters are masters at devising new methods of fabricating transactions thus requiring a consistent development and advancement of techniques for detecting and mitigating the falsifications. Numerous strategies have been proposed and used in the identification and mitigation of fraudulent transactions. Fraud detection and prevention involves analysis of the spending behavior of customers with the main aim being to uncover undesirable behaviour. It is focused on identification of suspicious events in an expeditious manner. Bayesian networks are suitable for circumstances where some data is already known and received data is partially unavailable or uncertain. The objective of utilizing Bayes rule is based on its ability to accurately predict the value of a selected discrete class variable given a set of attributes. The naïve Bayes technique is preferred due to its simplicity in dealing with training data and also its ability to handle missing values.
