Biosketch
Yurou earned her Bachelor’s degree in Mathematics in Business and Economics from the University of Mannheim, and later completed her Master’s degree in Mathematics with a concentration in Statistics and Applied Mathematics at the Technical University of Munich (TUM).
She is currently working towards her PhD under the guidance of Prof. Mathias Drton in the Mathematical Statistics group at TUM. Her primary research areas are Graphical Models and Causal Discovery.
relAI Research
Learning and Leveraging Causal Models
Causal discovery identifies cause-and-effect relationships in complex systems, with applications in biology, finance, and machine learning. Many machine learning models use all available features without understanding their causal roles, reducing robustness and interpretability. My research focuses on learning causal structures with graphical models, particularly through nonparametric methods that are broadly applicable in practice. Because causal graphs are acyclic, traditional approaches rely on difficult combinatorial searches. Recent continuous acyclicity constraints make causal discovery more scalable. We extended these ideas to kernel-based approaches and causal discovery under latent confounding. Our current research incorporates human prior knowledge into causal discovery, enabling experts and algorithms to iteratively refine causal structures for more reliable and realistic decision-making systems.
Publications
Liang, Yurou; Zadorozhnyi, Oleksandr; Drton, Mathias: Kernel-Based Differentiable Learning of Non-Parametric Directed Acyclic Graphical Models. Proceedings of Machine Learning Research, ML Research Press, 2024 (Proceedings of The 12th International Conference on Probabilistic Graphical Models), 253-272