Cuong Nguyen

Google Scholar Github Twitter LinkedIn Email

I am a Research Fellow at the Centre for Vision, Speech and Signal Processing, University of Surrey, working on human-AI cooperation in Professor Carneiro’s group.

Before that, I was a Research Associate at the School of Computer Science, The University of Adelaide from April 2022 to January 2024. I received my PhD in Computer Science from The University of Adelaide in March 2022, an MPhil in Electronic Engineering also from The University of Adelaide in January 2018, and a B.S. in Mechanical Engineering from Portland State University in June 2012.

Areas of study

My main research interest include probabilistic machine learning and graphical models. In particular, I prefer modelling how the given data is generated (e.g., through a graphical model), then perform parameter inference on such a modelling using the Expectation - Maximisation or variational inference coupled with some numerical optimisation techniques. I am also interested in statistical machine learning (e.g., PAC learning).

Publications

Please refer to my Google Scholar, or click the button below to retrieve the list of my publications from my ORCID.

Experience

  • Research Fellow   -   March 2024 - present
    Centre for Vision, Speech and Signal Processing
    University of Surrey, UK
    • research in human-AI cooperation: Current setting in human-AI cooperation requires (i) each sample must be annotated by all human experts, and (ii) each annotation must be associated with the identification of a human expert. Such requirements complicate the data collection, resulting in a more expensive and time-consuming process. My research is to relax each of the two assumptions to make the human-AI cooperation more practical, and
    • co-supervise two PhD students in human-AI cooperation and multirater learning.
  • Research Associate   -   April 2022 - January 2024
    School of Computer Science
    The University of Adelaide, Australia
    • research in the identifiability of noisy label learning: The noisy label learning problem with a single noisy label per sample is non-identifiable, meaning that there are infinite number of models that can generate the given dataset. My research is to find conditions to make the problem identifiable (i.e., having one unique solution). In short, for a C-way classification problem, at least (2C - 1) noisy labels per sample must be collected,
    • co-supervise a PhD student in the same topic about noisy label learning, and
    • teach the course COMP SCI 1104 - Challenges in Computer Science.

Education

  • PhD in Computer Science   -   April 2018 - March 2022
    The University of Adelaide, Australia
    Supervised by Professor Carneiro, Gustavo and Dr Do, Thanh-Toan to research in meta-learning:
    • probabilistic meta-learning: extends from learning a point estimate of meta-learning models to a distribution of meta-learning models by employing variational inference,
    • PAC-Bayes meta-learning: extends the PAC-Bayes bound of single task learning to multiple tasks in meta-learning,
    • task modelling for meta-learning: adapts the latent Dirichlet allocation to model tasks in meta-learning and measure their similarity, and
    • task-weighting in meta-learning with trajectory optimisation: adapts the trajectory optimisation to weight the contribution of each task when training a meta-learning model.
  • MPhil in Electronic Engineering   -   August 2015 - January 2018
    The University of Adelaide, Australia
    Research in micro-scaled vibration energy harvesting, focusing on electret-based approach.
  • BS in Mechanical Engineering   -   July 2010 - June 2012
    Portland State University, USA
    Awarded a scholarship from Intel Products Vietnam to study oversea.

Teaching

Courses taught at The University of Adelaide
Year Semester Course Level Role
2023 2 Grand Challenges in Computer Science Undergraduate Intructor
2022 2 Grand Challenges in Computer Science Undergraduate Intructor
2021 1 Foundations on Computer Science Master Tutor
2021 2 Foundations on Computer Science Master Tutor
Back to top