Programme
A four-year doctoral programme training the next generation of AI researchers.
The programme consists of taught modules in your first year, before spending years two to four working with your supervisor on your dedicated research project, which you will have identified upon application.
Year 1
Your first year will provide you with a foundation in Machine Learning and AI, and an in-depth understanding of the implications of its application to solve real-world problems.
You will be based at The University of Manchester, in one of the oldest Computer Science departments in the UK – the birthplace of the first stored-program computer and where Alan Turing first proposed the ‘Turing Test’.
Year 1 core modules
Machine learning is increasingly being used for decision support in data driven applications. A key concept when making decisions based on predictive models is that of uncertainty, such as applications of AI where safety or trustworthiness are required. Uncertainty quantification recognises that exact predictions are often out-of-reach due to theoretical or practical limitations. This course unit studies different probabilistic machine learning models that incorporate uncertain reasoning and the mathematical concepts and algorithms required to learn such models from data.
The module 'Machine learning and the physical world' is focused on machine learning systems that interact directly with the real world. Building artificial systems that interact with the physical world has significantly different challenges compared to the purely digital domain. In the real world, data is scarce, and often uncertain and decisions can have costly and irreversible consequences. However, we also have the benefit of centuries of scientific knowledge that we can draw from. This module will provide the methodological background to machine learning applied in this scenario. We will study how we can build models with a principled treatment of uncertainty, allowing us to leverage prior knowledge and provide decisions that can be interrogated.
Reinforcement learning (RL) looks to create machine learning models that can make decisions. An agent learns to achieve a goal in an uncertain, potentially complex environment. Successful real-world applications include but are not limited to robotics, control, operation research, games, economics, and human-computer interactions. This course will cover the breadth of modern model-free RL methods, discuss their limitations and introduce various current research topics. In particular, we expect to cover the following: deep learning methodology and architectures, stabilisation of approximated value estimation, modern actor-critic methods, planning as inference, exploration with deep networks, offline reinforcement learning, deep multi-agent reinforcement learning, multi-task and meta-learning.
This module provides an introductory level of training on the ethics in AI through three parts:
- Operationalising ethics in AI: focuses on incorporating ethical AI principles into the systems design process.
- Team coding hackathons: you’ll work together to create, innovate and hack on ideas that inspire you. You will brainstorm a project idea under a given thematic topic and present your idea through coding demonstrations. You'll learn necessary basic programming skills, including the use of Git, Docker, efficient Python programming and the use of Jax/Pytorch framework.
- Paper reproducibility challenge: this part investigates reproducibility of papers accepted for publication at top conferences, and verifies the empirical results and claims in the paper by reproducing the computational experiments, either via a new implementation or using code/data or other information provided by the authors. The students should submit their results in the form of blog posts.
You will undertake a research project, which will provide you with the skills to manage and deliver a technical project in the broad area of AI for decision making in complex systems. You also learn the communication skills needed to communicate the outcome of your work in a written report and presentation to your cohort.
Year 1 elective modules
You will also have the choice of two elective 15-credit modules, shown in the drop-down below.
Along with the core Year One modules, you will also have the choice of 2 elective 15-credit modules, which currently consist of the following, but please note these may be subject to availability:
- Principles of advanced engineering materials
- Advanced composites
- Cognitive robotics and computer vision
- Decision behaviour, analysis and support
- Design and analysis of randomised controlled trials
- Introduction to health data science
- Introduction to health informatics
- Foundations of machine learning
- Machine learning and advanced data methods
- Mathematical computing for medical imaging
- Mathematical programming and optimisation
- Multi-omics for precision medicine
- Fundamental mathematics and statistics for health data science
- Representation learning
- Risk, performance and decision analysis
- Simulation and risk analysis
- Statistical modelling and inference for health
- Tutorials in advanced statistics
Years 2 – 4
These three years will primarily focus on doctoral research, during which you will undertake a topic-specific research project working in close collaboration with your supervisor.
- You’ll be required to write an annual written report and viva at the end of each year of research.
- Alongside your PhD project, you’ll also have access to training modules in the Science and Engineering Doctoral Academy, which will help you build skills in academic writing, preparing presentations and personal development.
- To support your professional development, there are sessions to help you develop the skills needed in a diverse range of careers, such as media training, policy engagement and public engagement training. Several of our industry partners also offer placements for you to gain real-world experience.
- You’ll also be encouraged to participate in annual activities such as:
- Journal club: You’ll learn how to present highly advanced literature and disseminate research.
- Summer or winter schools: You’ll have the option to attend one external school per year.
- CDT conference: We will run a conference each year in which you can present your research and get feedback on your work. Industry partners will be invited, allowing you to network with industry professionals.