Postdoctoral Fellow – Semi‑Supervised Learning & Diffusion Models

About the role

The Wallenberg‑NTU Postdoctoral Fellowship invites outstanding early‑career researchers who earned their Ph.D. at a Swedish university to join Nanyang Technological University for a two‑year research appointment. The fellow will work on cutting‑edge semi‑supervised learning (SSL) methods, integrating entropy‑based regularisation and diffusion models for industrial data challenges.

Key responsibilities

• Conduct a comprehensive literature review of state‑of‑the‑art SSL techniques.
• Design and implement novel regularisation strategies and loss functions leveraging entropy principles.
• Adapt denoising diffusion probabilistic models (DDPMs) for manifold learning and data augmentation within the SSL pipeline.
• Develop rigorous benchmarking protocols using noisy industrial datasets (e.g., anomaly detection, predictive maintenance).
• Write clean, reproducible Python code (JAX, PyTorch, TensorFlow) and maintain technical documentation.
• Prepare research findings for submission to top‑tier machine‑learning conferences (NeurIPS, ICML, ICLR) and internal reports.

Required profile

• Ph.D. in Computer Science, Machine Learning, Statistics or a related quantitative field.
• Strong analytical skills with experience designing ablation studies.
• Proven ability to communicate complex mathematical concepts to interdisciplinary teams.
• Track record of implementing algorithms from academic papers and handling real‑world, sparsely‑labelled datasets.

Required skills

• Python programming.
• Experience with JAX, PyTorch or TensorFlow.
• High‑performance computing (HPC) clusters or cloud GPU instances.
• Deep‑learning model development and debugging.

What we offer

• Two‑year fixed‑term fellowship with access to NTU’s research facilities.
• Opportunities for interdisciplinary collaboration and high‑impact publications.
• Exposure to industrial datasets and real‑world problem solving.