A Review of Data-Driven Approaches with Emphasis on Machine Learning Base Intrusion Detection Algorithms

Abstract

The importance of the internet across the globe cannot be over-emphasized as such network security is essential to curb future attack occurrences. Cyber-attacks like DDoS and Ransomware yielded a lot of damage to connected devices by endangering and accessing them, notwithstanding these damages are air marked to be on the rise. To overcome these issues, machine learning has been used in different computing aspects such as cyber–Intrusion Detection. Recently, deep learning, extreme learning, and deep extreme learning networks have superseded machine learning in this context due to their iterative hidden layers that can manipulate complex features of cyber intrusion data. Hence, this research surveys the application of data-driven intelligent algorithms for cyber security attack detection in comparison to conventional machine learning techniques. The review focuses on the performance evaluation of several state-of-the-art intelligent algorithms and provides research gaps and future 2trends in the context of Data Security Attacks and Cyber Intrusion Detection.