An artificial neural network for bank robbery risk management: The OS.SI.F web on-line tool of the ABI Anti-crime Department

Abstract

The ABI (Associazione Bancaria Italiana) Anti-crime Department, OS.SI.F (Centro di Ricerca dell'ABI per la sicurezza Anticrimine) and the banking working group created an artificial neural network (ANN) for the Robbery Risk Management in Italian banking sector. The logic analysis model is based on the global Robbery Risk index of the single banking branch. The global index is composed by: the Exogenous Risk, related to the geographic area of the branch, and the Endogenous risk, connected to its specific variables. The implementation of a neural network for Robbery Risk management provides 5 advantages: (a) it represents, in a coherent way, the complexity of the "robbery" event; (b) the database that supports the AN is an exhaustive historical representation of Italian Robbery phenomenology; (c) the model represents the state of art of Risk Management; (d) the ANN guarantees the maximum level of flexibility, dynamism and adaptability; (e) it allows an effective integration between a solid calculation model and the common sense of the safety/security manager of the bank. © 2009 Springer-Verlag Berlin Heidelberg.