Phishing Website Detection with XGBoost and Adaptive Hyperparameter Optimization using the Bat Algorithm

Phishing websites are a growing threat to internet users, and traditional detection methods like blacklisting or relying on SSL certificates are no longer enough to keep up with the rapidly changing landscape of cyberattacks. In this study, we propose a new approach that combines the power of XGBoost, a popular machine learning algorithm, with the Bat Algorithm for adaptive hyperparameter optimization, specifically for detecting phishing websites. The Bat Algorithm, inspired by how Bats use echolocation, helps fine-tune critical hyperparameters like learning rate and maximum tree depth, making XGBoost more accurate and better at learning patterns in the data without overfitting. This approach strikes a balance between exploring new solutions and refining the best ones, leading to improved classification performance. Our experiments show that this method significantly enhances accuracy, achieving 94.27% across multiple datasets. Overall, this integrated approach offers an efficient and reliable solution for detecting phishing websites, providing a valuable tool in the ongoing fight against online threats and improving cybersecurity.