Currency Security Validation System Using Machine Learning and Deep Learning Algorithms

Abstract

Detection of counterfeit currency is one of the major challenges to which a financial institution faces and it needs to have an automated and precise system of checking the currency to protect transactions. This paper suggests a model of Currency Security validations based on data analysis of denomination, weight, size, color score, and security score to categorize the authenticity. It consists of machine learning and deep learning methods to determine performance on a heterogeneous sample of currencies. The experimental results are strong predictors with the Random Forest model having an accuracy of 96.80 and the neural network having 97.70. These results indicate the efficiency of AI-powered approaches to credible currency authentication.