Workshop: Predictive modelling in the context of uncertainty and heterogeneity in biomedical data

Tuesday 11 October 2022, 10:00 - 16:00

Introduction

Launched in June 2021, the Turing-Roche Strategic Partnership is a five-year collaboration between The Alan Turing Institute and Roche.

The partnership’s ‘North Star’ is to enable the generation of insights to better understand patient and disease heterogeneity and its relevance to clinical outcomes at an unprecedented level of precision in order to improve clinical care. 

Last year we ran a funding call and associated workshops on the theme of structured missingness, funding two 18-month projects. This year we are exploring the theme of predictive modelling in the face of uncertainty and heterogeneity in biomedical data, and inviting applications for funding.

Similar to last year’s successful model, we will be running a workshop for researchers to encourage discussion and exploration of this theme and have the opportunity to meet others to develop potential collaborative proposals for the call.

About the event

We are inviting researchers to an in-person workshop in London on Tuesday 11 October 2022, 10:00 - 16:00.

The workshop aims to convene a community of researchers interested in predictive modelling in the context of uncertainty and heterogeneity in biomedical data. Our intention is to bring a diverse group together, including established and early career researchers, to seek synergies and develop collaborative project ideas for the funding call.

We will be exploring three sub-themes during the workshop and having discussions in these areas in smaller groups, with a larger group discussion and debrief in the final part of the day. The three research sub-themes are:

  1. Generalisation despite heterogeneity: developing models that perform robustly in the face of data heterogeneity and generalise across patient heterogeneity
  2. Prediction uncertainty in personalised healthcare: quantifying the uncertainty of personalised (as opposed to generalised) predictions at the individual patient level made from models trained at the population level
  3. Algorithmic explainability and fairness for responsible AI: developing models that perform fairly across different patient subgroups and the decision process is interpretable and transparent 

We hope to work with attendees to produce a potential output from the workshop such as a report, perspective piece, or publication.

Lunch and refreshments will be included. There is limited travel funds available for those who may not have the funds available. Please contact the organisers if you have any questions. 

Apply to attend

Applications are now closed. You can find out more about the funding call here and other opportunities to engage with the partnership here.

Associated funding call

The funding call is inviting 12-18 months and shorter pilot proposals (~6 months) for each sub-theme.

All active attendees of the workshop will be eligible to apply for predictive modelling funding call and we encourage collaborative proposals that may form at the event e.g. that combine different areas of data science and/or different academic institutions.

The call will also be open to researchers who were not able to attend the workshop.

Key dates

Workshop applications open: Monday 5 September 2022

Workshop applications close: Monday 26 September 2022

Application update: Tuesday 27 September 2022 

Workshop takes place: Tuesday 11 October 2022 

Organisers