Daniel Neider

Daniel Neider is a Professor in the Department of Computer Science at TU Dortmund University, where he leads research on Verification and Formal Guarantees of Machine Learning. He is also affiliated with the Center for Trustworthy Data Science and Security at the University Alliance Ruhr and has previously held a research group leader position at the Max Planck Institute for Software Systems, Kaiserslautern. He teaches courses at both TU Dortmund and RPTU Kaiserslautern and is a principal investigator in multiple research projects.

Research Interests:
Daniel Neider's research lies at the intersection of machine learning and formal methods, with a focus on ensuring the safety, reliability, and trustworthiness of AI systems. His work combines symbolic reasoning from logic with inductive techniques from machine learning to develop automated tools for verification, synthesis, and explainability. Key areas include the verification of learning systems (e.g., robustness of neural networks), explainability of AI decisions, learning-based synthesis of reactive systems, specification learning, and the integration of automata learning into reinforcement learning. His theoretical work extends into automata theory, game theory, and logic.

Publication Trends:
His recent publications show a strong emphasis on neuro-symbolic methods, temporal logic inference, robustness certification of neural networks, and learning-based verification. He frequently publishes in top venues such as AAAI, IJCAI, TACAS, FMCAD, and CAV, reflecting his leadership in both AI and formal methods communities.

• Scientific Awards: No specific awards are mentioned in the provided text.

Advising and Grants:
Daniel Neider supervises numerous Master's and Bachelor's students and advises PhD candidates, particularly on topics combining formal methods with AI. He is a principal investigator in several funded projects, including the DFG-funded Temporal Logic Sketching and the DAAD-supported LeaRNNify, which explore specification assistance and the synergy between grammatical inference and neural network learning.

Labs and Teams:
He leads a research group focused on building practical tools for trustworthy AI. The group has developed influential software such as ICE, Horn-ICE, libalf, and QUGA, which are used for invariant synthesis, program verification, automata learning, and verification of deep autoencoders. These tools are publicly available on GitHub and Bitbucket, and some are accessible via web demos.