Private and Verifiable Machine Learning

The world is entering an extraordinary new era driven by advances in Artificial Intelligence (AI) and new emerging technologies (NextGen) and media dynamics.  We view the rapid improvements as a unique, once-in-a-generation opportunity to transform and achieve significant, ethical and lasting benefits for our country and the world.  

We are looking for an exceptional candidate to join our wider Secure Ethical AI and Future Digital Emerging Technologies interdisciplinary group working across computing and engineering and media working in response to the UK government’s AI and Digital agenda and the coming research NextGen trends. We will work to help International, National and local organisations to better understand, utilise and protect themselves and the general-public, and to help realise the benefits from the NextGen Technologies. 

The field of computing on encrypted data is gaining significant attention due to the scalability challenges current AI technologies face concerning privacy and verifiability. Key technological advancements such as Fully Homomorphic Encryption (FHE), Secure Multi-party Computation (MPC), Zero-Knowledge Proofs (specifically zkSNARKs and zkSTARKs), and Verifiable Computation offer promising solutions for various applications.  

These include but are not limited to: 

• Privacy-Preserving Machine Learning
• Privacy-Preserving Data Analytics
• Privacy-Preserving Deep Learning

However, different cryptographic schemes are better suited for different use cases. Therefore, based on application scenario an efficient switching mechanism between these schemes is of utmost important. The primary challenge lies in the high computational and communication costs associated with large computations on big data. This project aims to revisit the security module more formally and reduce these costs by optimizing the underlying cryptographic primitives and protocols. 

The research is biased in the School of Computing with a direct link to working with our Nationally recognised Cyber Security and CCSR teams. 

The PhD Programme 

The DMU PhD provides a solid education for a research career whether you stay in academic research or move to industry. There is a well-developed Researcher Development Programme which you will undertake alongside your research, supporting you through each year of your PhD. Working closely with your supervisors you will undertake open-ended research. As part of a vibrant PhD community in the Faculty you will have training opportunities which lead to presentations at international conferences and in writing papers in leading academic journals as the research progresses into the later stages of the PhD. The student will graduate with skills and knowledge at the for front of academic and industrial expectations. 

Starting date: There are three intakes a year at DMU and it is expected the successful candidate will start either in January 2025, April 2025 or October 2025. 

Funding Notes:  

This is a self-funded project and you will need to have sufficient funds in place (e.g., through a scholarship, personal funds and or other sources) to meet the tuition feeds and living expenses for the duration of the degree programme.  

DMU has information on scholarships and other funding opportunities at https://www.dmu.ac.uk/doctoral-college/study/scholarships.aspx 

How to apply: 

To apply for a research degree, you will need to apply via our Admissions Portal https://dmuhub.dmu.ac.uk/apply/index.html. You are required to provide the documentation as listed within section 5b at  https://www.dmu.ac.uk/doctoral-college/study/how-to-apply.aspx 

Studied at DMU previously? If you have studied at DMU previously, please submit the documentation listed in 5b via email to pgradmissions@dmu.ac.uk, not through the admissions portal. You will also need to complete and attach our application form and annexe form alongside the documents requested below: 

You will need to select a programme code when applying from a list of research degree programme codes. Please select the programme code closest to your research proposal.