Michele Guerra

Michele Guerra is a Doctoral Research Fellow at the Department of Mathematics and Statistics, UiT The Arctic University of Norway. His research focuses on graph neural networks (GNNs), with emphasis on explainability, temporal modeling, and subgraph-based architectures. Recent work includes applications of Koopman theory for GNN interpretation and probabilistic load forecasting using reservoir computing techniques.

Research Trends: His publications highlight innovative approaches to improving the expressivity and transparency of graph-based machine learning models, particularly through stochastic methods and subgraph decomposition. Key application domains include energy systems forecasting and dynamic graph analysis.

Key Collaborations: Active collaborations with researchers such as Indro Spinelli, Simone Scardapane, and Filippo Maria Bianchi are evident in recent co-authored works presented at venues like the Northern Lights Deep Learning Workshop.