FURI | Spring 2025

Analysis of Different Machine Learning Models to Detect Phishing Websites

Phishing attacks continue to evolve, allowing them to bypass traditional methods of detection and resulting in billions of dollars of losses. Existing machine learning algorithms have shown limited capabilities in adaptability and accuracy. This research aims to evaluate different machine learning algorithms and, through the use of hyperparameter tuning, improve baseline models. Performance will be evaluated using F1 scores, comparing the different models to select the one with the highest accuracy. Enhanced phishing detection capabilities can significantly reduce financial fraud and bolster cybersecurity resilience.