Abstract: The issue of finance fraud is on the rise and poses a significant threat to the financial industry. While various techniques have been developed to tackle this problem, data analysis is proving to be a particularly effective approach. By analyzing vast amounts of complex data from finance databases, data analysis can help automate the process of detecting fraudulent activities. This approach has already been successfully employed in detecting credit card fraud during online transactions. However, credit card fraud detection is a challenging problem due to two main factors: the profiles of normal and fraudulent behaviors change frequently, and credit card fraud datasets are highly skewed. To address this challenge, this project proposes investigating and evaluating the performance of various algorithms on highly skewed credit card fraud data. The dataset used in this project contains 284,786 transactions from European cardholders. The algorithms will be applied to both raw and preprocessed data, and their performance will be evaluated based on accuracy, sensitivity, specificity, and precision.

| DOI: 10.17148/IARJSET.2023.10838