Financial Fraud Detection with Improved Neural Arithmetic Logic Units

Domain specific neural network architectures have shown to improve the performance of various machine learning tasks by large margin. Financial fraud detection is such an application domain where mathematical relationships are inherently present in the data. However, this domain hasn’t attracted much attention for deep learning and the design of specific neural network architectures yet. In this work, we propose a neural network architecture which incorporates recently proposed Improved Neural Arithmetic Logic Units. These units are capable of modelling mathematical relationships implicitly within a neural network. Further, inspired by a real-world credit payment application, we construct a synthetic benchmark dataset, which reflects the problem setting of automatically capturing such mathematical relations within the data. Our novel network architecture is evaluated on two real-world and two synthetic financial fraud datasets for different network parameters. We compare our proposed model with several well-established classification approaches. The results show that the proposed model is able to improve the performance of neural networks. Further, the proposed model performs among the best approaches for each dataset.

[1]  Nitesh V. Chawla,et al.  SMOTE: Synthetic Minority Over-sampling Technique , 2002, J. Artif. Intell. Res..

[2]  Chris Dyer,et al.  Neural Arithmetic Logic Units , 2018, NeurIPS.

[3]  Helmut Krcmar,et al.  Reducing false positives in fraud detection: Combining the red flag approach with process mining , 2018, Int. J. Account. Inf. Syst..

[4]  Xipeng Qiu,et al.  Neural Arithmetic Expression Calculator , 2018, ArXiv.

[5]  Stefan Axelsson,et al.  Paysim: a financial mobile money simulator for fraud detection , 2016 .

[6]  Aihua Shen,et al.  Application of Classification Models on Credit Card Fraud Detection , 2007, 2007 International Conference on Service Systems and Service Management.

[7]  David J. Hand,et al.  Statistical fraud detection: A review , 2002 .

[8]  VARUN CHANDOLA,et al.  Anomaly detection: A survey , 2009, CSUR.

[9]  Karlis Freivalds,et al.  Improving the Neural GPU Architecture for Algorithm Learning , 2017, ArXiv.

[10]  Lukasz Kaiser,et al.  Neural GPUs Learn Algorithms , 2015, ICLR.

[11]  F. Wilcoxon Individual Comparisons by Ranking Methods , 1945 .

[12]  Andreas Dengel,et al.  Detection of Anomalies in Large Scale Accounting Data using Deep Autoencoder Networks , 2017, ArXiv.

[13]  Andreas Hotho,et al.  iNALU: Improved Neural Arithmetic Logic Unit , 2020, Frontiers in Artificial Intelligence.

[14]  Raghavendra Chalapathy University of Sydney,et al.  Deep Learning for Anomaly Detection: A Survey , 2019, ArXiv.

[15]  Changjun Jiang,et al.  Credit Card Fraud Detection Using Capsule Network , 2018, 2018 IEEE International Conference on Systems, Man, and Cybernetics (SMC).

[16]  Alexander Rosenberg Johansen,et al.  Measuring Arithmetic Extrapolation Performance , 2019, NeurIPS 2019.

[17]  Reid A. Johnson,et al.  Calibrating Probability with Undersampling for Unbalanced Classification , 2015, 2015 IEEE Symposium Series on Computational Intelligence.

[18]  Alex Graves,et al.  Grid Long Short-Term Memory , 2015, ICLR.

[19]  Miklos A. Vasarhelyi,et al.  Predicting credit card delinquencies: An application of deep neural networks , 2018, Intell. Syst. Account. Finance Manag..