A Study on the Efficacy of Machine Learning Models in Intrusion Detection Systems

Abstract

The electronics industry has seen a rise in demand for faster and more affordable delivery due to developments in information technology. Technology is developing quickly, which simplifies living but also presents a number of security issues. As the Internet has grown over time, so too have the amount of online attacks. The intrusion detection system (IDS) is one of the supporting layers that can be utilized for information security. IDS avoids questionable network activity and provides a pristine environment for conducting business. In the process of building an e-commerce system, the most challenging aspect is ensuring user security during online transactions. Security methods for intrusion detection were investigated in this study. The need for ongoing intrusion detection monitoring stems from the need for continued technological adaptation, which leads to a comparison of adaptive artificial intelligence-based intrusion detection systems. This paper demonstrates the use of reinforcement learning (RL) and regression learning-based intrusion detection systems (IDS) to very challenging problems, including resource allocation and input feature selection.