In the era of digitalization, utilization of credit cards is on the rise for acquiring goods through both online and offline avenues. Fraudulent credit card transactions have been on rise nowadays. This study used machine learning algorithms to detect fraudulent activities. The model was trained using machine learning algorithms logistic regression, random forest and xgboost. Implemented was carried out using HTML and CSS as a web application. The algorithms are compared and the one with greatest accuracy, precision, recall, and F1-score is considered the best algorithm for fraud prediction. Our findings indicates that XGboost has the highest accuracy of 99.86%, followed by Random Forest Classification with 99.84% and Logistic Regression with 99.41%.Random forest classification and XG Boost models demonstrated good performance in predicting fraudulent transactions, while the logistic regression model performed poorly in this regard. These results offer insight to target users about the performance of three different fraud detection models.