Learning to Optimize for Data Science
Not stated
- Funding
- Competition Funded PhD Project (Students Worldwide)
- Application deadline
- Year-round applications
About the project
About the Project Optimization algorithms are fundamental to modern data science and machine learning, underpinning applications from training AI models to solving large-scale inverse problems. However, traditional optimization methods are usually based on fixed update rules and often require substantial manual tuning. This PhD project will investigate Learning to Optimize (L2O) , where machine learning is used to automatically design or improve optimization algorithms. The project will explore how components such as step sizes, momentum, preconditioners, update directions, proximal operators, or optimization geometries can be learned from data. A particular focus will be on developing structured and interpretable learned optimizers that combine the flexibility of machine learning with the reliability of classical optimization [1,2]. The project will study both the theoretical properties of learned optimization algorithms, such as convergence, stability, and generalization, and their practical performance on applications arising in machine learning, computational imaging, generative modelling, and statistical inference. The project is suitable for students with a strong background in applied mathematics and statistics (or computer science), and an interest in optimization and machine learning. If you are interested in this project, please email your CV and transcript to Dr Junqi Tang ( j.tang.2@bham.ac.uk ), and please also remember to state clearly whether you are willing to be self-funded if no funding is available or applicable to you. Funding Notes: For UK and EU candidates: Funding may be available through a college or EPSRC scholarship in competition with all other PhD applications: https://www.birmingham.ac.uk/research/centres-institutes/research-in-mathematics/mathematics-phd-information Strong candidates are encouraged to make an informal inquiry. For non-UK/non-EU candidates: Strong self-funded applicants will be considered. For Chinese candidates: The China Scholarship Council (CSC) Scholarship: https://www.csc.edu.cn/chuguo China Scholarship Council (CSC) PhD Scholarships Programme at the University of Birmingham PhD Placements and Supervisor Mobility Grants China-UK: https://www.britishcouncil.cn/en/programmes/education/higher/opportunities/phd