ERCRFS: Ensemble of Random Committee and Random Forest using StackingC for Phishing Classification

Abstract

The process of gathering sensitive information such as credit card/ debit card number, password, OTP etc., through a deceptive e-mail or SMS with links to Malicious site is called Phishing. An attacker during the Phishing attack masquerades as a trustworthy entity of some kind and lures the victim to provide his/her sensitive information. As per the latest data breach reports, the Phishing attacks make up almost a third of all data breaches. The attackers are using more sophisticated tools and are getting away with all the important data . Our work examines the application of Machine Learning Algorithms for the efficient classification of Phishing. The MMR, MSDMR and FSF Algorithms contribute towards the selection of significant Features from the data set such that there is no degradation in the performance of Classification. The hybrid framework ERCRFS, proposed by us ensures better Detection Accuracy values amongst all existing models.