Florian Felten

Florian Felten is a postdoctoral researcher at ETH Zürich's Department of Mechanical and Process Engineering, affiliated with the Artificial Intelligence in Engineering Design group. His research focuses on developing algorithms for automated design using reinforcement learning and multi-objective optimization. He holds a PhD from the University of Luxembourg (excellent thesis award) and an M.Sc. in Computer Science from Université Catholique de Louvain.

• Education:
• PhD in Computer Science, University of Luxembourg (2021)
• M.Sc. in Computer Science, Université Catholique de Louvain, Belgium (2018)

His research interests include multi-objective reinforcement learning (MORL), decomposition-based methods, and software frameworks for benchmarking. He contributed to the development of tools like MO-Gym, MORL-Baselines, and a comprehensive taxonomy for MORL/D frameworks. His work emphasizes practical applications in engineering design and optimization.

Key achievements include a 2023 NeurIPS publication on benchmarking tools and a 2024 Journal of AI Research paper establishing a MORL taxonomy. He has also explored metaheuristics for exploration strategies in MORL and presented work at OLA and ICAART conferences.

• Awards:
• Excellent Thesis Award for PhD work on Multi-Objective Reinforcement Learning

Felten's research bridges reinforcement learning and optimization, with applications in automated design and algorithm development. He collaborates on open-source libraries and benchmarks to advance the field.