A fraud detection model that applies accounts’ behavior features and outlier detection methods are proposed. Given a set of transactions, the accounts are grouped into two groups according to the location that the transactions took place, i.e., ‘local-only’ and ‘has-abroad’. A feature is extracted to reflect the normal behavior of the accounts in each group. Only known legitimate transactions are used to extract a set of features for representing a legitimate behavior. An unknown transaction is classified either normal or fraud using an outlier detection. The experimental result shows that the proposed feature with an Isolation Forest outlier detection technique is able to detect all fraud transactions.
CITATION STYLE
Laimek, R., Kaothanthong, N., & Supnithi, T. (2018). ATM Fraud Detection Using Outlier Detection. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 11314 LNCS, pp. 539–547). Springer Verlag. https://doi.org/10.1007/978-3-030-03493-1_56
Mendeley helps you to discover research relevant for your work.