Paper Title
Enhancing Credit Card Fraud Detection with a Hybrid Unsupervised Learning Approach

Abstract
With their seininternetandon line shopping, credit cardsarenow more used for purchases,therebyincreasingonline financialfrauds.Since2023,creditcardfraudshaveresulted in a loss globally of around $32.39 billion USD. These fraudshave victimized many innocent people, mainly deterred older adults, less mindful of technology, from carrying out online payment activities. This would imply that, traditionally, detection of credit card fraud relied highly on supervised machine learning approaches that require the existence of the labeled datasets and antecedent knowledge related to the fraud transaction patterns. Currently, more efforts are made to apply unsupervised learning for credit card fraudetection by improving the performance of techniques such as k-means clustering. These techniques are frequently combined with optimization algorithms like Particle Swarm Optimization (PSO), Grey Wolf Optimization (GWO), Ant Colony Optimization (ACO), and Genetic Algorithms (GA). The work unveiled different strategies of optimization that have been found to be effective in enhancing unsupervised learning models in the detection of credit card fraud. Keywords – Particle Swarm Optimization(PSO),Grey Wolf Optimization (GWO), Ant Colony Optimization (ACO), Genetic Algorithm (GA),K-means Clustering