Probabilistic Deep Learning Group

The Probabilistic Deep Learning (PDL) Group, led by Prof. Dr. Magda Gregorová, explores the theoretical foundations of modern machine learning, with emphasis on probabilistic modeling and deep learning. Our research focuses on understanding and advancing modern deep generative models from a rigorous mathematical perspective. We are particularly interested in unsupervised and semi-supervised learning methods that extract meaningful structures from data. Through theoretical analysis and algorithmic innovation, we aim to develop principled approaches that push the boundaries of what deep learning systems can learn and represent.