A Review of Nature-Inspired Artificial Intelligence and Machine Learning Methods for Cybersecurity Applications

Abstract

Listen

Translate article

Machine learning (ML) and artificial intelligence (AI) methods are some of the latest advancements in the field of computing. Among these methods, there are nature-inspired techniques such as deep learning and deep neural networks, which are inspired from the neural networks of the human brain. These methods are applicable towards the security of networks and network-connected machines from malware, intrusion, and other cyberattacks. For example, in dealing with modern cyber threats, some standard ML and AI methods that can be useful are malicious code recognition for malware analysis, object-based modeling to classify security threats, and heuristic rule systems for intrusion detection. In this way, ML and AI can play a key role in cyber threat detection and prevention. Due to the large amounts of data packets passing through a network, processing and parsing through that data to find malware, intrusion, or other malicious code and files can be overwhelmingly difficult for humans. Machine learning models can be trained to detect malicious patterns in data or files and can thus be used to automatically detect malware or intrusive activity. Additionally, humans are limited in terms of the amount of time or duration that they can spend, but once programmed, a machine learning model can continue running and operating nonstop to detect and prevent malicious code and files from entering a network-connected system. This can reduce human effort and minimize human error by automating the computing required to detect and thwart cyberattacks. This paper surveys and reviews different AI and ML methods that have been used in past literature for cybersecurity applications. The goal of this work is to aid cybersecurity researchers and professionals on how to employ AI and ML techniques for cybersecurity applications, such as malicious code detection, intrusion detection, and cyber threat analysis.