Schmidt
Wiss. Mitarbeiter/-in
M.Sc. Johann Schmidt
Institut für Intelligente Kooperierende Systeme (IKS)
AG Artificial Intelligence Lab
AG Artificial Intelligence Lab
Universitätsplatz 2, Gebäude 29,
39106 Magdeburg,
G29-022
vCard
Education
- 2014-2018 Bachelor Electrical Engineering, University of applied Science Magdeburg-Stendal, Germany
- 2018-2020 Master Digital Engineering, Otto-von-Guericke University Magdeburg, Germany
- since 2020: Researcher / PhD student at the AI-Lab of the Otto-von-Guericke University Magdeburg, Germany
Main Research Interests
- Find solutions to Combinatorical Optimization problems (as job scheduling, TSP, etc.) using Deep Learning
- Traffic Light Control (TLC) with Deep Reinforcement Learning and Graph Neural Networks
- Learn to structure latent spaces by leveraging symmetries in the data domain (Geometric Deep Learning).
Funded Projects
- During SENECA (2020-2022) we developed a self-learning decision support system for real-time job sequence and machine allocation planning.
- In PASCAL (2022-2025) we build a proactive smart controller for traffic light control in Magdeburg.
- AIEngineering (2022-2026) fuses engineering with artifiicial intelligence forming a new practise-oriented Bachelor Study Program.
Follow me
Publications
- J. Schmidt and S. Stober. Tilt your Head: Spatial Transformation Invariance via Input Canonicalization during Inference. International Conference on Machine Learning (ICML), 2024. [PDF]
- J. Perschewski, J. Schmidt, S. Stober. Pursuing the Perfect Projection: A Projection Pursuit Framework for Deep Learning. International Workshop on Self-Organizing Maps and Learning Vector Quantization, Clustering and Data Visualization (WSOM+), 2024.
- J. Schmidt, B. Köhler and Hagen Borstell. Reviving Simulated Annealing: Lifting its Degeneracies for Real-Time Job Scheduling. The Hawaii International Conference on System Sciences (HICSS), 2024. [PDF]
- J. Schmidt and S. Stober. Learning Continuous Rotation Canonicalization with Radial Beam Sampling. ArXiv PrePrint, 2022. [PDF]
- S. Lang, T. Reggelin, J. Schmidt, M. Müller, and A. Nahhas. NeuroEvolution of augmenting topologies for solving a two-stage hybrid flow shop scheduling problem: A comparison of different solution strategies. Expert Systems with Applications, 2021. [PDF]
- J. Schmidt and S. Stober. Approaching Scheduling Problems via a Deep Hybrid Greedy Model and Supervised Learning. IFAC Symposium on Information Control Problems in Manufacturing, 2021. [PDF]
- S. Bexten, J.Schmidt, C. Walter, and N. Elkmann. Human Action Recognition as part of a Natural Machine Operation Framework. International Conference on Emerging Technologies and Factory Automation (ETFA), 2021. [PDF]