Hidden Markov Model Using Transaction Patterns for ATM Card Fraud Detection

• Activity Number: 130 - Statistical Computation, Simulation, and Computer Experiments
• Type: Contributed
• Date/Time: Monday, July 31, 2017 : 8:30 AM to 10:20 AM
• Sponsor: Section on Physical and Engineering Sciences
• Author(s): Alfred Akinsete* and Bridget Nkemnole
• Companies: Marshall University and University of Lagos
• Keywords: Hidden Markov Model ; Generalized Poisson ; Optimal Detection ; Forward-Backward Algorithm ; Estimation

Abstract

Estimating the rate at which transactions happen in the social and business world is becoming imperative and has been attracting a lot of research attention. The increase in transactions using the ATM card has become bedeviled with increasing rate of attendant fraud associated with it. This study proposes a hidden Markov model (HMM) based on the Poisson distribution (HMM[Pois]), the generalized Poisson distribution (HMM[GenPois]), and the Gaussian distribution (HMM[Gauss]) for which optimal detection of patterns of anomalies is computed using the forward-backward algorithm. The proposed estimation procedure based upon the three distributions for the HMM model is used to construct a sequence of operations in ATM card transaction processing, and detect fraud by studying spending profile of the cardholder, followed by checking an incoming transaction against spending behavior of the cardholder. If the transaction satisfies a predefined threshold value, then the transaction is decided to be legitimate else, the transaction is declared as fraudulent. The evaluation statistics used shows that the HMM[Gauss] is the most appropriate model in detecting ATM card fraudulent transactions.