Title :
Application and Efficiency of Probabilistic Neural Network in Suspicious Transaction Surveillance
Abstract :
We apply the PNN neural network to suspicious transaction surveillance by deviation of time series data prediction. Comparing with other type neural networks including BP, Elman, LVQ, and SOM, we find that PNN has obvious advantages in structure complexity and efficiencies which are suitable in analysis for massive financial data.
Keywords :
Bayes methods; Biological neural networks; Classification algorithms; Neurons; Probabilistic logic; Training; Probabilistic Neural Network; outlier detection; suspicious financial transaction;
Conference_Titel :
Measuring Technology and Mechatronics Automation (ICMTMA), 2015 Seventh International Conference on
Conference_Location :
Nanchang, China
Print_ISBN :
978-1-4673-7142-1
DOI :
10.1109/ICMTMA.2015.50
