A Survey Current Datasets used for Intrusion Detection using Machine Learning

Abstract

Cyberattacks in today's digital age cause the loss of sensitive data and a huge financial loss for enterprises and countries. Therefore, the role of the cyber security expert is very important to protect data from increased and new attacks. Researchers focus on anomaly-based intrusion detection systems to detect these unknown attacks and machine learning algorithms play a vital role in this process because they detect attacks accurately. Data sets currently used in intrusion detection systems suffer from a clear lack of real network threats, attack representation, and include a large number of abandoned threats, which limit the accuracy of detection within the current intrusion detection systems' methods of machine learning, which make them unable to trace increasing and new attacks in cloud environments, containers. This research paper aims to combine classification and analysis of existing data sets in order to improve the creation new data sets that simulate the actual reality of the network's real data. This will improve the efficiency of the next generation of intrusion detection systems and reflect network threats more accurately