Our ability to design intelligent machines that learn models directly from data has led to tremendous progress across a plethora of scientific fields in recent years. Examples range from communications and signal processing to medical imaging, genetics, finance and social sciences. The objective of our research group at Imperial College London is to develop algorithms that enable provable and efficient learning across a variety of challenging environments.
We invite applications for a fully-funded PhD studentship at the intersection of optimisation, machine learning and network science. Projects can focus on any of the following areas:• Distributed, federated, or decentralised learning for multi-agent systems
• Multi-task and meta-learning
• Robust, privacy- or communication-constrained learning
• Optimisation for machine learning (e.g., nonconvex landscapes, optimal algorithms)
• Information-theoretic concepts for (distributed) learning
The studentship covers funding for tuition (including international fees) as well as a stipend. Applicants will need to satisfy general entry requirements and apply through the departmental portal (https://www.imperial.ac.uk/electrical-engineering/study/phd/), listing Dr Stefan Vlaski as potential supervisor. Additionally, applicants are expected to demonstrate strong aptitude for analytical reasoning, excellent communication skills, and creativity.Applicants wishing to discuss details of their application are invited to email Dr Stefan Vlaski directly (s.vlaski@imperial.ac.uk). Details on research are also available online (https://stefanvlaski.github.io).
