Protocol based Ensemble Classifier for Malicious URL Detection
Himanshu Kumar, Prabhav Gupta, Rajendra Prasad Mahapatra · 2018
The paper presents a study of ensemble based machine learning approach for URL classification. Effective prediction of URL as malicious or benign is imperative requiring timely detection. Already existing models for classification were exhaustive using the same feature list for URL following either HTTPS or HTTP protocol. Ensemble classifiers are presented along with the procedure discussed in detail for selecting a robust feature list. The model was trained on five ensemble based algorithms. Results obtained show that boosted trees outperforms all the other ensemble based algorithms giving highest accuracy for both protocols.