Top of page

Student/Staff Portal
Global Site Navigation

School of Science

Local Section Navigation
You are here: Main Content

Related Contents

Muhammad Umer

Overview of thesis

Android remains the world’s most widely used mobile operating system, making it a prime target for malware developers. Android malware continues to evolve in sophistication, using techniques such as code obfuscation, dynamic payload loading, and permission misuse to evade traditional signature-based detection systems. As a result, machine learning (ML) and deep learning (DL) approaches have been widely explored for Android malware detection. While many existing studies report high detection accuracy, several practical gaps remain. First, many recent approaches rely on computationally intensive deep learning models, which are difficult to deploy on mobile devices or at the edge. Second, many models act as “black boxes”, offering limited explainability, which reduces trust and practical adoption in security operations. Third, a number of studies rely on outdated or limited datasets, raising concerns about the real-world applicability and the potential for concept drift in malware behaviour. This research aims to address these gaps by investigating lightweight and explainable machine learning techniques for Android malware detection, with a focus on practical deployment and contemporary datasets.

Contact

Muhammad Umer
PhD Student
Master of Science by Research (Cybersecurity)
School of Science
Skip to top of page