Miltiadis Kofinas

Miltiadis (Miltos) Kofinas is a Postdoctoral Researcher at the Climate Extremes Group within the Institute of Environmental Studies (IVM) at Vrije Universiteit Amsterdam. He holds a Diploma (MSc equivalent) in Electrical and Computer Engineering from Aristotle University of Thessaloniki and is completing his PhD at the University of Amsterdam's Video & Image Sense Lab under Prof. Efstratios Gavves. His work focuses on AI methods for climate science, particularly foundation models for weather forecasting, and has explored applications in autonomous vehicles and geometric deep learning techniques.

Research interests include graph neural networks, neural fields, parameter-space networks, and equivariant representations. He has collaborated on projects involving explainable AI, hybrid dynamical systems, and benchmarking neural field training methods. His academic journey includes a Diploma thesis on Scene Graph Generation using GNNs at Aristotle University and industry experience at P.A.N.D.O.R.A. Robotics as a computer vision engineer.

Key research trends in his publications span AI-driven climate modeling, geometric deep learning innovations, and improving interpretability of complex AI systems. His work often bridges theory with practical applications in environmental science and autonomous systems. Active collaborations focus on interdisciplinary challenges in AI and environmental studies.

Currently affiliated with VU Amsterdam's Climate Extremes Group, his research integrates advanced machine learning techniques to address climate-related forecasting challenges. He has contributed to foundational work in neural field optimization, hybrid system modeling, and equivariant architectures that respect physical symmetries.