Can artificial neural networks detect fraud? A case study
AbstractArtificial neural networks have been widely applied in many various areas, e.g. optimization, pattern recognition or time series forecasting. They also have been used in classification tasks, for instance to state whether a potential debtor is going to be solvent or, perhaps, in a determined time it is highly probable that they will default and go bankrupt. The efficiency of neural network in that task is similar to other methods like Discriminant Analysis or credit rating methods. However, the question arises if, at that initial moment, it is possible to define the possibility of potential fraud. The paper investigates if neural networks can be of any use in prior fraud detection. The problem is presented as specific classification task basing on historical data. Various topologies of neural networks are used. Results of those different methods are juxtaposed, and their performance compared. The study can be classified in applied studies group and the research strategy is descriptive.
|Publication size in sheets||0.5|
|Book||Szkutnik Włodzimierz, Sączewska-Piotrowska Anna, Hadaś-Dyduch Monika, Acedański Jan (eds.): 9th International Scientific Conference: Analysis of International Relations 2018. Methods and Models of Regional Development. Winter Edition. Conference Proceedings, 2018, Wydawnictwo Uniwersytetu Ekonomicznego w Katowicach, ISBN 978-83-7875-421-3, [978-83-7875-420-6], 198 p.|
|Keywords in Polish||sieć neuronowa, oszustwo|
|Keywords in English||credit risk, default, neural networks|
|Score||= 20.0, 24-06-2020, ChapterFromConference|
* presented citation count is obtained through Internet information analysis and it is close to the number calculated by the Publish or Perish system.