About this role
Statistical methods are leveraged in many scientific applications to: specify a data collection practice (experimental design), draw conclusions from sparse and noisy data (inference), and assess uncertainty about those conclusions (uncertainty quantification). Bayesian inference and experimental design methods are increasingly used in scientific practice, and offer appealing theoretical guarantees when certain assumptions hold. However, whether these assumptions hold is difficult or impossible to verify in practice.
The successful candidate will undertake a project in the area of statistics/data science with the goal of (i) understanding the consequences of violations of core assumptions on the behaviour of Bayesian methods, and/or (ii) developing methods to mitigate these consequences. There will be a particular emphasis on applications from psychology, cognitive science and computer science, and the candidate will be encouraged to engage in interdisciplinary collaboration and communication.
We are looking for candidates with a background in or demonstrated ability to learn about: Bayesian methods, probabilistic machine learning or inverse modelling. Prospective applicants are encouraged to direct informal inquiries to Dr. Sabina Sloman ([email protected]).
Funding notes:
The scholarship will cover home tuition fees, training support, and a stipend at standard rates for 3-3.5 years. Self-funded students worldwide are welcome to apply.