Analysis of Credit Card Fraud Transaction Detection using Machine Learning Algorithms
Shashank Sahu, Neeta Sahu · 2023
Credit cards are now used in daily life. The need for online business transactions is rising as quickly as the internet. Markets are expanding thanks to the internet, which influences customers to buy the things they need. The inability to raise enough funds can occasionally prevent you from getting the things you need. For a limited time, businesses are also offering good discounts on their products. It forces the user to need to buy things. Use of credit card is necessary to satisfy the user needs. Additionally, credit card companies provide numerous discounts and reward points for using credit cards to make purchases. When the number of people increases in using credit cards, then the chance of fraudulent transaction is also increases. Finding out about fraudulent credit card transactions becomes urgently necessary. Unauthorised and illegal actions are seriously hindering the expansion of the credit card industry. Financial losses result from fraud in many circumstances. Credit cards are utilised on websites. Numerous methods are offered by machine learning algorithms which can be used for determining that whether a transaction is fraud or not. This study compares various machine learning algorithms and demonstrates that the KNN method performs best, with a 99.9% accuracy rate. This provides a thorough review of the effectiveness of various machine learning methods, which is helpful in creating applications for detecting credit card fraud. This paper also shows that lowest accuracy of 98.6% is achieved using Naive Bayes algorithm.