Variational decentralised learning
Candidates are expected to have solid foundations in statistics, information theory, optimisation methods, and demonstrated research outputs (quality publications) and programming skills.
The PhD study aims to originate variational decentralised learning theories and methods to integrate variational, decentralised and deep learning to satisfy complex stylistic, local-global integrative requirements.
Key details
- 20268140, 20268141, 20268142
- PhD
- Applications close 1 March 2026 for international students, 30 April for domestic
- Domestic, International
- Information technologies
- $39,700 p.a.
The scholarship will sponsor foundational research on variational decentralised learning, including
- characterising distributional discrepancy and vulnerability
- quantifying decentralised non-IIDnesses
- modelling stylistic uncertainties
- transforming existing averaging-aggregation learning frameworks.
Availability
This scholarship is available to eligible candidates to undertake a direct entry three-year PhD program.
Components
The scholarship comprises:
- a tuition fee offset/scholarship
- a living allowance stipend.
The value of each stipend scholarship is $39,700 per annum (full time, indexed) for three years.
How to apply
Before submitting your application, submit an expression of interest (EOI) to Professor Longbing Cao at longbing.cao@mq.edu.au.
Your EOI should include:
- selected publications
- your CV
- other evidence of research foundations and achievements.
Candidates without solid foundations and demonstrated research outputs will not be eligible for the scholarship.