Microeconomics: machine learning model with behavioural intelligence to reduce credit card fraud
Debasis Chakraborty, Abhijit Paul, Gurdeep Kaur
Abstract
Credit card fraud - is the adversity complementary to a modern facile payment system. Over the years, credit card frauds have managed to cause massive losses to banks and credit companies globally. Where the extensive availability of data was envisaged to improve customer service, the fraudsters have managed to exploit this boon to amplify the volume, intensity, and value of frauds. While a complex and continually developing problem is hard to convey successfully using rule-based algorithms, machine-learning models present themselves as adept means of tackling the crisis. By combining supervised and unsupervised models, incorporating behavioural analytics and domain knowledge, and using artificial neural networks, fraudulent transactions detection is achievable with desirable accuracy and poses to be a promising tool for diminishing considerable losses.