4-year PhD Studentship: Machine learning for understanding causes of disease

This project will explore using machine learning to predict health outcomes using data in epidemiology cohorts (e.g. UK Biobank https://www.ukbiobank.ac.uk/), and the use of explainable AI approaches to identify potential new predictors of disease. For example, you might choose to explore using deep learning – the current state of the art of machine learning prediction approaches – to build models predicting health outcomes with large and complex datasets [1].

The particular data you work with will depend on your particular area of interest, but examples are predicting health outcomes from physical activity data measured with accelerometers [2] or imaging data [3].

Aims and objectives

Aim: To identify novel determinants of disease outcomes using cutting-edge machine learning approaches.

Objectives:

1. To develop machine learning models (e.g., using deep learning) to predict health outcomes using large-scale health data (e.g., from UK Biobank).
2. To investigate the potential to understand the predictions using explainable AI approaches.

There may also be opportunity to follow-up identified associations to look at causal relationships with Mendelian randomization [4].

Methodology

You will develop expertise in using machine learning for making predictions in health (e.g., using deep learning approaches), and explainable AI approaches to understand predictive models. You will also develop knowledge in epidemiology and working with health data e.g., data wrangling, high performance computing.

How to apply for this project

This project will be based in Bristol Medical School - Population Health Sciences in the Faculty of Health Sciences at the University of Bristol.

Please visit the Faculty of Health Sciences website for details of how to apply