PhD Candidate in Machine Learning for Temporal and Spatiotemporal Data

The Department of Computer Science (IDI) at NTNU in Trondheim is seeking a PhD Candidate for a four-year position focused on machine learning for temporal and spatiotemporal data. This role is part of the interdisciplinary STRIDE project, which investigates the resilience of transport systems against disruptions, shocks, and crises.

The Research Focus

You will develop deep learning methods to forecast reliably under distribution shifts and detect disruptions within complex networks. Potential research directions include:

• Developing time-series foundation models, including pretraining on synthetic data.
• Creating spatiotemporal graph neural networks to model disruption propagation.
• Implementing anomaly detection and localization for time series on graphs.

Responsibilities

• Complete doctoral education and conduct high-quality research.
• Develop, implement, and evaluate deep learning methods, including open-source code and benchmarks.
• Disseminate findings through academic publications and popular science.
• Collaborate within the DART research group and the interdisciplinary STRIDE project team.
• Contribute to teaching and career-promoting activities at the department (totaling one year).

What We Offer

NTNU provides a strong international research environment with access to high-performance computing resources and GPUs. You will benefit from a mentor program, career guidance, and an inclusive work culture. The position includes a gross salary of approximately NOK 580,000 per annum and membership in the Norwegian Public Service Pension Fund.