Hybrid Machine Learning Model to Detect and Mitigate EMAIL Phishing Attacks
Akbarsagari Mohammad Fazil, Pallapolu Sai Vardhan Reddy, Jaini Eswar, S. Prabakeran · 2025
The research develops an improved email classification system through the hybridization of Support Vector Machines with XGBoost to fight expanding phishing attacks while using TF-IDF vectorization for selecting significant text features. The classification pipeline of the classifier starts with SVM classifying examples from which XGBoost receives scores to perform refined classification. Multiple layer classification systems improve detection performance and enhance both the precision-recall measurements. A new signature extraction method enables the system to use TF-IDF for keeping and analyzing the key phishing patterns needed for real-time monitoring. The lightweight adaptive solution excels above conventional methods because it handles changing phishing threats effectively.