Abstract
Phishing attacks continue to pose a significant threat to individuals and organizations, making it crucial to develop effective countermeasures. Machine learning algorithms have shown promise in detecting and mitigating phishing attacks. The study evaluates the performance of four popular algorithms in the context of phishing detection and compares the effectiveness of these four different algorithms; Random Forest, Decision Tree, XGBoost, and Logistic Regression, to determine which one achieves the highest accuracy. The results show that XGBoost outperforms the other algorithms and can accurately detect phishing attacks with a high degree of precision. The algorithms are compared based on factors such as training time, test time, model size, interpretability, and explainability. To compare the effectiveness of these algorithms, the study conducted experiments using a dataset of phishing emails. The algorithms were trained on a labeled dataset and evaluated based on metrics such as accuracy, precision, and recall. The results demonstrate that XGBoost outperforms the other algorithms, achieving the highest accuracy in detecting phishing attacks. The findings of this study have significant implications for the development of antiphishing technologies. By leveraging machine learning algorithms, particularly XGBoost, organizations can enhance their ability to detect and prevent phishing attacks. This can help protect individuals' personal information, passwords, and credit card numbers from falling into the hands of cybercriminals.