Information Leakage Control in Machine Learning Models for Privacy Assurance

These projects are open to students worldwide, but have no funding attached. Therefore, the successful applicant will be expected to fund tuition fees at the relevant level (home or international) and any applicable additional research costs. Please consider this before applying. 

In the contemporary landscape of artificial intelligence (AI), the growing use of machine learning models raises concerns about the inadvertent leakage of sensitive information. This comprehensive PhD project is designed to address the intricacies of information leakage in machine learning models by examining attack models, exploring, quantifying, and mitigating risks while considering ethical and legal dimensions.

The project initiates with an exploration of contemporary attack models and methodologies used for quantifying information leakage in machine learning models. By synthesising existing literature and dissecting prevalent attack strategies, the research will pinpoint gaps and opportunities for enhancement. This exploration serves as a critical starting point, providing insights into the vulnerabilities that contribute to unintended information leakage. Next, building upon the insights gained, the research will propose and design an innovative framework dedicated to quantifying the extent of information leakage by leveraging game-theoretic approach and differential privacy. Third, to fortify privacy assurances in machine learning models, the project will develop novel mitigation strategies that enhance privacy without compromising the models' performance, aiming to obtain an optimal balance applicable to diverse real-world scenarios. In parallel, the research will meticulously examine the ethical implications associated with information leakage in AI models. This holistic consideration will extend to proposing guidelines that account for the societal impacts of AI. Finally, the proposed techniques and frameworks, informed by both attack models and quantification methods, will undergo rigorous experimental validation. Diverse datasets representing various application domains will be used, addressing challenges related to scalability, generalisation, and adaptability of the proposed methods.

Computing Science at University of Aberdeen is globally recognised for its excellence in Artificial Intelligence. As part of our research team, the candidate will have access to state-of-the-art research facilities and work with a highly collaborative and interdisciplinary team of researchers from different fields including agent, machine learning, cyber security, formal method, natural language processing. The candidate will also have opportunities to present their research findings at top-tier conferences and publish in high-impact academic journals.

Essential Background:

Decisions will be based on academic merit. The successful applicant should have, or expect to obtain, a UK Honours Degree at 2.1 (or equivalent) in Computer Science, Mathematics, Cyber Security, or a related field along with:

• Interest in privacy-preserving technologies.
• Strong programming skills (Python, TensorFlow and, PyTorch …).
• Familiarity with C/C++, various scripting languages, and the Linux environment.
• Ability to work collaboratively and independently on research projects

Application Procedure:

Formal applications can be completed online: https://www.abdn.ac.uk/pgap/login.php.

You should apply for Computing Science (PhD) to ensure your application is passed to the correct team for processing.

Please clearly note the name of the lead supervisor and project title on the application form. If you do not include these details, it may not be considered for the studentship.

Your application must include: A personal statement, an up-to-date copy of your academic CV, and clear copies of your educational certificates and transcripts.

Please note: you DO NOT need to provide a research proposal with this application.

If you require any additional assistance in submitting your application or have any queries about the application process, please don't hesitate to contact us at [Email Address Removed]