Johannes Czech

Johannes Czech is a Researcher in the Machine Learning Group at the Computer Science Department of Technische Universität Darmstadt. His research focuses on applying deep learning models to planning algorithms like Monte-Carlo Tree Search (MCTS), with a particular emphasis on parallel reinforcement learning and supervised learning optimizations. He has co-supervised numerous theses on topics such as phase-specific learning in Pommerman, neural network architectures in AlphaZero, and evaluating MCTS variants for chess and other games.

Education:

• Ph.D. student in Computer Science (2020–present), TU Darmstadt
• M.Sc. in Computer Science (Visual Computing, 2017–2020), TU Darmstadt
• B.Sc. in Computer Science (2014–2017), Hochschule Furtwangen University

Research interests include optimizing MCTS for imperfect information games, improving feature representations in AlphaZero, and exploring generative adversarial networks for creative applications like emoji generation. His work bridges theoretical algorithm design with practical implementations in domains like chess variants and robotics.

Teaching: Co-instructor for courses such as "Einführung in die Künstliche Intelligenz" (Introduction to AI) and supervised projects like LiGround, an open-source Chess Variant Analysis GUI. He has also contributed to research projects involving time management in chess using neural networks and human data.

Lab affiliations: Machine Learning Lab at TU Darmstadt, collaborating with researchers like Prof. Dr. Kristian Kersting and Dr. Arturo Crespo.