Skip to main content
School of Electronic Engineering and Computer Science

Efficient and Safe Learning within Reinforcement Learning Systems

Supervisor: Dr David Mguni

Project Description:

The project aims to develop a new set of techniques to induce high sample efficiency and safe learning within Reinforcement Learning. A key focus area is to explore how methods from Game Theory, Bilevel Optimisation, and Optimal Control Theory can offer new alternative techniques to induce greater sample efficiency and enable reinforcement learning methods to better satisfy safety constraints during training.

PhD topics can explore any of the following or combinations thereof:

o Transfer learning within Reinforcement Learning and Hierarchical Reinforcement Learning towards the goal of efficient training.

o PAC-Bayes theory within (multi-agent) Reinforcement Learning.

o Supervised learning and model estimation techniques within model-based Reinforcement Learning.

Apart from the above, I am also open to supervising projects that aim to develop multi-agent tools for application within foundation models.

Prerequisites:

- A Master’s degree (Distinction or equivalent) or an expected completion of such qualifications before starting the PhD.

- A keen interest in Reinforcement Learning (RL) and Multi-agent Reinforcement Learning (MARL).

- Strong mathematical skills e.g., linear algebra, calculus, real analysis etc.

- Proficiency in programming within python and standard ML packages (e.g. TensorFlow, Pytorch, etc.). Experience of using RL in simulated environments.

The PhD studentship is funded by EPSRC Doctoral Landscape Award open to those with Home and International fee status. However, the number of students with international fee status who can be recruited is capped according to the EPSRC terms and conditions, so competition for international places is particularly strong. The PhD student will receive an annual stipend of £21,237 for the academic year 2024/25, with funding available for a duration of up to 3.5 years.

How to apply

Queen Mary is interested in developing the next generation of outstanding researchers and decided to invest in specific research areas.

Applicants should work with their prospective supervisor and submit their application following the instructions at: http://eecs.qmul.ac.uk/phd/how-to-apply/.

The application should include the following:

- CV (max 2 pages)

- Cover letter (max 4,500 characters) stating clearly in the first page whether you are eligible for a scholarship as a UK resident (https://epsrc.ukri.org/skills/students/guidance-on-epsrc-studentships/eligibility)

- Research proposal (max 500 words)

- 2 References

- Certificate of English Language (for students whose first language is not English)

- Other Certificates

Please note that in order to qualify as a home student for the purpose of the scholarships, a student must have no restrictions on how long they can stay in the UK and have been ordinarily resident in the UK for at least 3 years prior to the start of the studentship. For more information please see here.

Application Deadline

The deadline for applications is the 29th January 2025.

For general enquiries contact Mrs. Melissa Yeo (administrative enquiries) or Dr. Arkaitz Zubiaga (academic enquiries) with the subject “EECS 2025 PhD scholarships enquiry”.

For specific enquiries contact Dr. David Mguni.

Back to top