News and Events

Symposium – Data Science in Public Health 

 

Description

Please follow this link to register for the symposium.

The UCL Centre for Data Science is hosting a half-day symposium on Data Science in Public Health on Wednesday 3 December 2025, at De Morgan House, London Mathematical Society.

UCL has a rich and diverse community of researchers applying data-driven methods to address public health challenges — from digital health and epidemiology to policy evaluation and health equity. This symposium will bring together researchers from across the university to showcase recent work and foster cross-disciplinary discussion on the opportunities and challenges of using data science to improve population health.

Full Programme

Session 1 (13:10 - 14:15)

• Nathan Green (Statistical Science)
  Estimating the Determinants of Health Literacy for Policy Prioritisation: A Local Level Case Study in Newham, London
• Gabrielle Marsden (Wolfram Research)
  Data with a Pulse: Real-World Insights for the NHS

Session 2: (14:35 - 15:40)

• Vincent Grigori (Population, Policy & Practice)
  Virus Watch: Understanding community incidence, symptom profiles, and transmission of COVID-19 in relation to population movement and behaviour
• Rob Saunders (Clinical, Educational & Health Psychology)
  Data Science and Mental Health: Possibilities and Challenges

Panel Discussion (16:00 - 17:00)
Panellists:

• Emma Beard (Epidemiology & Public Health)
• Christina Pagel (Clinical Operational Research Unit)
• Laura Horsfall (Institute of Health Informatics)
• Steve Harris

Talk abstracts

Estimating the Determinants of Health Literacy for Policy Prioritisation: A Local Level Case Study in Newham, London

Dr Nathan Green

This talk will present a methodological framework to identify and quantify the local-level determinants of health literacy, focused on Newham, London. The analysis combines heterogeneous data sources: the Newham Residents Survey (NRS), the UK 2021 Census, the ONS Skills for Life (SfL) survey, and others. Health literacy is defined using literacy, numeracy, and Information and Communication Technology (ICT) metrics. To address data sparsity and missingness, we employed iterative preprocessing and imputation. A Multilevel Regression with Post-stratification (MRP) model is used to generate conditional estimates of health literacy. We then adopt an Average Treatment Effects (ATE) approach to quantify the strength of association for key determinants. Model robustness is assessed through sensitivity analyses against other survey datasets. To translate these findings for local decision-makers, we also derived a scoring and ranking framework inspired by concepts from network meta-analysis. The resulting data-driven tool quantifies determinant importance with uncertainty, aiding the prioritization of resources for targeted health literacy interventions.

Data with a Pulse: Real-World Insights for the NHS

Dr Gabrielle Marsden

The NHS holds one of the world’s richest sources of real-world health data — but unlocking its full potential requires a clear awareness of the data and the system’s limitations, an understanding of clinical priorities and powerful computational tools. In this talk, I’ll share practical applications of Wolfram technology to transform NHS data into actionable insights that inform real decisions across the health system. From identifying opportunities to invest in diabetes prevention, to analysing workforce gaps in haematology, improving patient safety in hospitals, and supporting GPs in managing patients with long-term conditions, I’ll demonstrate how a computational approach can bridge the gap between data and care. The session will highlight the real-world impact of this work and offer a glimpse of how data-driven insight can help shape a healthier, more efficient NHS.

Virus Watch: Understanding community incidence, symptom profiles, and transmission of COVID-19 in relation to population movement and behaviour

Dr Vincent Grigori

Virus Watch is an award-winning community cohort study of COVID-19, with around 60,000 participants across England and Wales. Running June 2020 until April 2025, Virus Watch brings together multiple disciplines to tackle the pandemic including epidemiology, data science, immunology, geospatial science, and psychology amongst others. The Virus Watch data has been used to produce over 30 peer-reviewed publications receiving extensive digital and broadcast media coverage. Virus Watch work has also been presented to the Scientific Advisory Group for Emergency (SAGE) contributing to the evidence base for and implementation of policies such as the introduction of booster dose vaccines. Furthermore, research using the Virus Watch data has been used as evidence in the UK COVID-19 Inquiry, particularly the collaboration between Virus Watch, the Race Equality Foundation and Doctors of the World, which demonstrated the inequalities experienced by some groups during the pandemic. Internationally, Virus Watch publications have been selected as part of the United States Centre of Disease Control’s Public Health Genomics and Precision Health Knowledge database. The Virus Watch work has also been cited by the Australian Technical Advisory Group on Immunisation.

Data Science and Mental Health: Possibilities and Challenges

Prof Rob Saunders

Mental health conditions, such as depression and anxiety disorders are increasing in prevalence globally, with far-reaching impacts across individuals, their relationships, and wider society. Whilst evidence-based treatments exist, many individuals are unable to access care, and a significant proportion do not respond to interventions provided. Data science has a valuable role in identifying mental health inequalities, as well as providing insights into potential solutions to improve individual outcomes. This talk will discuss the challenges of data science in mental health, including issues with measurement and sources of data, as well as examples operationalising data science to generate insights in the field. These include examples leveraging routinely collected data in the NHS Talking Therapies for anxiety and depression programme (previously known as IAPT). With over 1.8 million referrals to these services each year, such data provides a valuable opportunity to understand how to better support mental health at scale.

 

 

 

 

Seminar Series: 

Date	Seminar

 

Previous Events: 

Syposium on Computational Statistics and Machine Learning in the Sciences 

 

The UCL Centre for Data Science is organizing a one-day symposium on computational statistics and machine learning (CSML) on the 5th December 2024. The symposium will be held at De Morgan House in the London Mathematical Society. 

UCL has a very wide breadth of research in CSML spread across most of its faculties. This symposium aims to bring these researchers together to discuss recent advances in terms of novel algorithms, theory or applications in the sciences and engineering.

Confirmed speakers include Sam Livingstone (Statistical Science) Harita Dellaporta (soon-to-be Statistical Science), Jason McEwen (Mullard Space Science Laboratory), Marta Betcke (Computer Science), Ziheng Yang (Genetics, Evolution and Environment) and Niall Jeffrey (Physics & Astronomy).

The symposium will start at 1.30pm with a series of invited talks, and will conclude from 6-7pm with a reception allowing attendees from all across UCL to get to know one another.

Symposium on Causal Inference 

The UCL Centre for Data Science is organizing a one-day symposium on causal inference on 22 November. The symposium will be held at De Morgan House in the London Mathematical Society.

This symposium aims to bring together different researchers at UCL who work on or are interested in the different aspects of the theory and application of causal inference. It will also include external leaders in the field as keynote speakers, as well as members of the industry that make use of causal models and methods.

The symposium will start at 10:30am with a reception followed by a session on the application of causal inference at UCL. After breaking for lunch, it will reconvene at 2pm for the keynote talk and a second session on causal inference in machine learning. These sessions will be followed by a panel addressing the challenges and opportunities of using causal models. The closing reception will take place at 5:30pm.

The confirmed list of speakers is as follows: Professor Philip Dawid (University of Cambridge), Professor Silvia Chiappa (Google DeepMind and UCL), Professor Neil Davies, Dr Ben Deaner, Professor Karla Diaz-Ordaz, and Professor Arthur Gretton (UCL and Google DeepMind).