Based at the UiT the Arctic University of Norway in Tromsø, our group activity is dedicated to basic research in machine learning. We also apply our research to a variety of domains, including energy analytics and climate science.

We’re excited to launch this blog as a platform to share our journey, insights, and breakthroughs with the broader community. Alongside the formal presentation in our papers, we aim to complement it here with a format that’s more relaxed, visual, and hands-on. In addition, we plan to publish tutorials that guide readers through key methods and concepts that we use in our work, making it more approachable to both researchers and practitioners.

Our Research Focus

Our work spans several exciting areas:

• 🎱 Graph Pooling: Techniques for down-sampling graph structures while preserving important information (library, AB25, CB25, GZB+22);
• 📊 Spatiotemporal modeling: Using graphs to model complex temporal dependencies (HCB25, MAB24, CMB+23);
• 🎯 Uncertainty Quantification: Assessing and mitigating uncertainty in forecasting in both structured and unstructured data domains (NCB+25, CJM+25, GSB23);
• 🔬 Interpretability: Making spatiotemporal models more transparent and trustworthy (GSB24, GSS+23);
• ⚡ Scalability: Breaking barriers to handle massive real-world datasets (NCB+25, CMB+23).

What You’ll Find Here

In this blog, we’ll be sharing:

• Research insights from our latest publications;
• Educational content to make complex concepts accessible;
• Practical applications of graph neural networks;
• Updates on our group activities, collaborations and events.
• Behind the scenes looks at our research process and team dynamics.

Stay tuned for our upcoming posts, where we’ll dive deep into our latest research.

The NGMLGroup Team
Northernmost Graph Machine Learning Group
UiT the Arctic University of Norway