Fully Funded PhD Studentship: Bayesian Modeling of High-Dimensional Structural Data
Fully FundedManchester, United Kingdomphd· £21,805 per year (tax-free, UKRI rate for 2026/27)
Project Overview
This project focuses on developing a comprehensive Bayesian learning framework for high-dimensional structural data with practical applications.
Basic Information - Full Project Title: PhD Studentship: Bayesian Modeling of High-dimensional Structural Data - Host Institution: The University of Manchester - Department: Department of Mathematics - Supervisor: Dr Nilabja Guha - Study Level: PhD - Duration: 3.5 years - Start Date: October 2026 - Funding Type: Fully Funded - Stipend: £21,805 per year (tax-free, UKRI rate for 2026/27)
- Application Deadline: 28 October 2026 (all year round)
Project Description
In many applications such as biological sciences, social science, and engineering, we encounter high-dimensional observations. Bayesian approach can provide a flexible modeling framework for underlying structures in high dimension such as underlying covariance structure, conditional dependency graphs etc. With the change in data-generating mechanism, these high-dimensional structures may change with time, where the change can depend on latent factors or variables. These projects will focus on developing a comprehensive Bayesian learning framework for this broad class of problems while focusing on specific applications. The goal would be to develop computationally efficient and scalable Bayesian learning methodologies with practical applications and establish relevant theoretical properties.
Requirements
- At least a 2.1 honours degree or a Master's (or international equivalent) in a relevant science or engineering-related discipline
How to Apply
Contact Dr Nilabja Guha (nilabja.guha@manchester.ac.uk). Please include details of your current level of study, academic background, any relevant experience, and a paragraph about your motivation to study this PhD project. Apply early as the advert may be removed before the deadline.
