Add Your Heading Text Here
Add Your Heading Text Here
Add Your Heading Text Here
Add Your Heading Text Here
Add Your Heading Text Here
Add Your Heading Text Here
Add Your Heading Text Here

Faithful Onwuegbuche

Student

Project Title

Machine Learning Techniques for Adaptive Ransomware Intrusion Detection

Project Description

This research explores machine learning-based techniques for adaptive intrusion detection of ransomware attacks. Ransomware is malicious software used to block access to computer systems, encrypt, exfiltrate, or damage files unless a ransom is paid. Ransomware is a severe cyber threat, causing significant losses to critical sectors like healthcare, education, finance, and infrastructure. Notable examples include the WannaCry attack on the UK’s National Health System in 2017 and the Irish health system attack in 2021costing £92m and €80m in direct costs and lost output respectively.
The rise of ransomware can be attributed to the financial benefits gained using cryptocurrencies as a means of payment, estimated to reach $265 billion globally by 2031. The COVID-19 pandemic has also contributed to the problem, as the shift to remote work has led some individuals to adopt inadequate security practices, resulting in a 600% increase in cybercrime. Additionally, the popularity of ransomware-as-a-service has allowed even novice attackers to launch sophisticated ransomware attacks.
Traditionally, signature-based methods have been used in detection, but these can be easily evaded by generating new variants and using obfuscation techniques. To overcome these limitations, machine learning has emerged as a prominent method for the detection and classification of ransomware. However, there are current issues in the literature on ransomware detection and classification using machine learning that this study seeks to address such as the lack of accurate, explainable, and adaptive detection models to keep pace with the constant evolution of ransomware, the lack of standard datasets, proper feature investigation, and the threat of adversarial machine learning attacks.
By tackling these issues, the research seeks to develop effective, adaptive, and explainable machine-learning techniques for robust ransomware intrusion detection that can effectively mitigate the rapidly evolving ransomware attacks.