معرفی
Dmitry Kobak is a group leader in the Department of Data Science at the Hertie AI Institute, University of Tübingen, Germany. He holds the title of Privatdozent at the Faculty of Computer Science and served as a visiting professor (Vertretungsprofessor) at Heidelberg University during the 2023/24 winter semester. His research focuses on machine learning and data science applications in biology, including self-supervised learning, dimensionality reduction, and topological data analysis. He is also engaged in statistical forensics, analyzing electoral fraud, war fatalities, and excess mortality patterns.
Education: BSc in Computer Science (St. Petersburg ITMO University), MSc in Theoretical Physics (St. Petersburg State University), PhD in Computational Motor Control (Imperial College London). Postdoctoral work with the Machens Lab (Champalimaud Institute) and Mehring Lab (Freiburg University/Imperial College London).
Teaching: Introductory machine learning courses for MSc students in Tübingen and BSc students in Heidelberg. Recent courses include Einführung ins Machinelle Lernen (German) and Transformers, Large Language Models, and their use in Physics (English).
Research supervision includes postdocs (Sebastian Damrich), PhD students (Rita González Márquez, Niklas Böhm), and multiple MSc students. Active in reviewing for top venues like NeurIPS, ICML, and Nature journals.
Labs/Teams: Member of the ELLIS Society, Cluster of Excellence «Machine Learning for Science», and IMPRS-IS associated scientist. His work bridges machine learning theory with practical applications in neuroscience, forensics, and biomedical research.
