PhD studentships (from September 2024 or any date thereafter, 4 years)

If you are interested in a PhD that brings together health-related topics and machine learning / artificial intelligence methods, please get in touch. If you are successful in your application, you will be based at the Centre for Artificial Intelligence (AI) in the Department of Computer Science at University College London (UCL). UCL is one of the top universities in the world and its AI centre one of the best places for this kind of research training in the UK.

PhD topics of interest can include (but are not limited to) the development of:

• machine learning methods for forecasting the prevalence of infectious diseases (influenza, COVID-19, and others) by incorporating online user-generated data streams (e.g. web search activity) and mechanistic approaches (our current work in this area),
• early warning systems for seasonal epidemics or novel disease pandemics using non-traditional information (e.g. online user activity); if successful, these systems may be incorporated into more conventional epidemiological practices nationally (UKHSA) or internationally (WHO), similarly to flu detector or our COVID-19 models,
• transfer learning models for monitoring the prevalence of diseases (across different locations and/or across different diseases) from online user-generated data streams (see [1, 2] for some proof-of-concept methods),
• machine learning methods for predicting the risk of a serious illness for individuals based on their online web search history (data collection from a patient cohort is ongoing), and
• language models for understanding health-related statements on social media platforms to facilitate their use in various downstream tasks.

The topics above can be broken down to PhD projects with a more theoretical focus, e.g. (a) the derivation of neural network architectures for time series forecasting, (b) the incorporation of uncertainty estimates in these models, (c) the joint optimisation of neural network forecasters with compartmental components, (d) transfer learning models, and (e) language models. Importantly, these PhD topics are not a definitive list, and I would like to encourage you to propose your own topic of interest that satisfies two requirements: having a health-related focus and be driven by machine learning.

Prior to applying formally, please send an email to v.lampos@ucl.ac.uk with:

• a CV (2 pages max), and
• optionally a research statement (6 pages max, excluding references) with a lay summary (abstract) of your proposed research, a background review of related work in this area, the core research proposal and how it fits within the current and past research landscape, and a description of what you would like to achieve (and how) in the first 6-12 months of your PhD. Please use this Overleaf template for your research statement.

Please note that although you can contact me without sending a proposal in the first instance, the development of a proposal will be an essential and very crucial part of your application process.

It will be beneficial to begin this process as early as possible in the academic year, e.g. November of year x-1, if you want to start a PhD in September of year x. Detailed instructions for formally applying are available here (general PhD in Computer Science call) and here (UCL research excellence scholarship, deadline: 12 January 2024, positions are competitive, but fully funded). Please note that applications unless otherwise stated are formally being reviewed twice in the academic year with soft submission deadlines in mid-January and mid-April, and that all PhD candidates will need to go through an interview process. For exceptional cases we might be able to expedite the process outside of these predetermined dates.

Research fellow (postdoc) positions

Please watch this space as a 2-year postdoc position co-supervised by me will be announced in early 2024 (funded by EP/X031276/1). Do not hesisate to get in touch if you want to find out more.