Aplikasi Machine Learning Untuk Deteksi Serangan Code Injection

Abstract

Code Injection is one of the top forms of cybersecurity attack in the modern world. To overcome the limitations of conventional signature-based detection techniques, and to complement them, several Machine Learning approaches have been proposed. When analysing this approach, the survey focused primarily on common intrusion detection, which can be further applied to specific vulnerabilities. In addition, between the steps of Machine Learning, there is a data pre-processing phase, which is very important in the data analysis process. This data pre-processing appears to have been the least researched in the context of Network Intrusion detection, i.e. in Code Injection. The aim of this survey is to fill the gap by analysing and classifying existing Machine Learning techniques, which are applied to Code Injection attack detection, with particular attention to Deep Learning . Our analysis reveals that the way input data is pre-processed greatly affects performance and attack detection rates. The proposed pre-processing cycle demonstrates how various Machine Learning -based approaches to detect Code Injection attacks take advantage of different input data pre-processing techniques.