Please use this identifier to cite or link to this item: https://hdl.handle.net/10419/336300 
Year of Publication: 
2020
Citation: 
[Journal:] Latin American Journal of Central Banking (LAJCB) [ISSN:] 2666-1438 [Volume:] 1 [Issue:] 1/4 [Article No.:] 100011 [Year:] 2020 [Pages:] 1-14
Publisher: 
Elsevier, Amsterdam
Abstract: 
We present a general supervised machine-learning methodology to represent the payment behavior of financial institutions starting from a database of transactions in the Colombian large-value payment system. The methodology learns a feedforward artificial neural network parameterization to represent the payment patterns through 113 features corresponding to financial institutions' contribution to payments, funding habits, payment timing, payment concentration, centrality in the payment network, and systemic effects due to failure to pay. We then use the representation to compare the coherence of out-of-sample payment patterns of the same institution to its characteristic patterns. The performance is remarkable, with an out-of-sample classification error around three percent. The performance is robust to reductions in the number of features by unsupervised feature selection. In addition, we confirm that network centrality and systemic effect features definitively contribute to enhancing the performance of the methodology. For financial authorities, this is a major step towards the automated detection of individual financial institutions' anomalous behaviors in payment systems.
Subjects: 
Feature selection
Machine learning
Neural networks
Pattern recognition
Payments
JEL: 
C45
E42
G21
Persistent Identifier of the first edition: 
Creative Commons License: 
cc-by-nc-nd Logo
Document Type: 
Article

Files in This Item:
File
Size





Items in EconStor are protected by copyright, with all rights reserved, unless otherwise indicated.