Two funded PhD projects in Bayesian/Machine Learning Methods are available through the MRC LID (London Intercollegiate Doctoral Training Partnership), for entry Autumn 2020:
Bayesian and probabilistic machine learning approaches for imputing and integrating spatiotemporal mobile phone location data (Supervisors Dr Alex Lewin and Prof Sanjay Kinra):
This project will build Bayesian and probabilistic machine learning models to integrate large-scale mobile phone location data in combination with geocoded built environment data to build improved estimates of environmental exposures to be used as risk factors in chronic disease studies. The project will use large-scale GPS location data from mobile phones, downloaded every few minutes over a period of several weeks, to be analysed in one integrated model. Bayesian and Probabilistic machine learning methods will be used to integrate the data using spatiotemporal models, including imputation of location data over missing time periods. The data comes from the APCAPS (Andhra Pradesh and Children Parents Study) cohort located in 29 villages 1-2 hours away from Hyderabad city, capital of Telangana state in South India. The improved exposure estimates will be used to assess the effects of diet and environment on cardiovascular disease risk factors and outcomes.
Bayesian and probabilistic machine learning models for predicting sub-clinical and clinical morbidity in electronic health record data (Supervisors Dr Alex Lewin and Prof Dorothea Nitsch):
The project will use big data (electronic health record data and UK biobank data, to be analysed in one integrated model), in order to predict the presence of undiagnosed or sub-clinical disease. This project will develop probabilistic models that can be applied in a range of scenarios (different diseases and different co-morbidities) in order to predict probabilities of a given person having an undiagnosed condition, based on their existing EHR data. The models will be developed in the context of specific conditions (heart failure, atrial fibrillation, COPD, proteinuria), which have reliable diagnostic data in UK Biobank.
LSHTM is a world-leading centre for research and postgraduate education in public and global health, with 3,900 students and more than 1,000 staff working in over 100 countries. The PhD projects will be based in the Faculty of Epidemiology and Public Health, across the Departments of Medical Statistics and Non-communicable Disease, research leaders in epidemiology and biostatistical methodology. The Doctoral Training Partnership provides excellent training infrastructure: a cohort of students, generous funding for training and travel and an extensive modular MSc including courses in advanced quantitative methods.
Both projects are suitable for candidates with a strong background in mathematics/statistics/computer science, with high interest in developing and applying methods to real life health data science applications.
Projects are funded through the MRC LID (UK/EU applicants), see here to apply:
https://mrc-lid.lshtm.ac.uk
Deadline is end of December 2019.
Please contact Dr Alex Lewin ([log in to unmask]) for informal discussion and further information about the projects.
You may leave the list at any time by sending the command
SIGNOFF allstat
to [log in to unmask], leaving the subject line blank.
|