Loading content...
The research proposes an AE-ProbRF approach for ATM card fraud prevention. PSO selects significant features, an autoencoder creates low-dimensional representations, and probabilistic Random Forest classifies authentic and fraudulent transactions. The model achieves 96.34% accuracy.
ATMs provide continuous banking services but are exposed to card skimming, cash trapping, physical tampering, theft and unauthorised access. The paper motivates intelligent automated surveillance and secure authentication mechanisms.