A Hybrid Machine Learning Model to Enhance Cybersecurity: An Integration of KNN, RF and XGBoost

Abstract

Living in the digital age has changed the way we do the majority of things, including living, working, and interacting, and positively impacted living in unprecedented ways when it comes to convenience, connectivity, and accessibility. Today, in the digital age, cybersecurity is the major concern. Cyber-attacks have become more sophisticated and frequent, there is a need for developing advanced approaches to protect computer systems and networks from being compromised. In recent years, machine learning has become one of the most powerful tools that can help to bring improvement in the cybersecurity field. However, current machine learning models aren't able to detect sophisticated cyber-attacks, like zero day and targeted attacks. In this paper, we investigate the limitations of current ML models in the field of Cybersecurity, and propose a framework for their improvement. The proposed framework includes the hybrid machine learning method to detect cyber-attacks. Performance evaluation has been done on KDD dataset than experiments show that the proposed hybrid model is superior to conventional machine learning model in terms of accuracy, precision, recall rate and F1-measure. Overall attack detection rates are improved by the integration. Overall proposed approach solutions to strengthen cybersecurity appear promising using machine learning. It will assist organizations to detect and prevent cyber-attacks better to secure their computer systems and networks.