Multi-Objective Reinforcement Learning
Florian Felten, ‘Multi-Objective Reinforcement Learning’, in Unilu - Université du Luxembourg [FSTM], Luxembourg.
CV
My LinkedIn is probably more complete than this page.
Education
- Ph.D in Computer Science (2021–2024)
- T.A. in Optimization for Computer Scientist: Master
- M.S. in Computer Science, Université Catholique de Louvain, Major in AI, Big Data and Optimization, 2018
- Thesis subject: “Reducing train delays in a real-time context: a Constraint Programming approach using conditional time-interval variables”
- Grade: Magna Cum Laude
- Erasmus at KTH, Stockholm
- B.S. in Computer Science, Université Catholique de Louvain, 2015
Work experience
- 2024–: PostDoc @ ETH Zurich
- 2023–: Project Manager @ Farama Foundation
- I’m one of the original developers and maintainers of MO-Gymnasium.
- I also contribute to other projects of the foundation (PettingZoo, SuperSuit).
- 2021–2023: Lecturer @ Uni.lu
- Programming Fundamentals 3
- 2nd year CS course
- Functional Programming, Concurrent Programming, Introduction to distributed systems
- 2018–2021: Software engineer @ N-SIDE
- Duties included: Architecture, Development and deployment of the apps, Backlog management, Requirement analysis.
- Stack: Scala, Akka, Kubernetes, IBM CP Optimizer, Scala.js
Skills
- Optimization
- Machine Learning
- Programming
- Functional
- Imperative & Object-Oriented
- Concurrency
- Distributed systems
- Teaching
- Chilling
Reviews
- Neural Computing and Applications (Reviewer)
- Multi-Objective Decision Making Workshop (MODeM) 2023 & 2024 (PC member)
- Adaptive and Learning Agents Workshop at AAMAS (ALA) 2023 & 2024 (PC member)
- International Conference on Optimization and Learning (OLA) 2022–2024 (PC member)
Publications
Multi-Objective Reinforcement Learning Based on Decomposition: A Taxonomy and Framework
Florian Felten, El Ghazali Talbi, and Grégoire Danoy, ‘Multi-Objective Reinforcement Learning Based on Decomposition: A Taxonomy and Framework’, in Journal of Artificial Intelligence Research.
A Toolkit for Reliable Benchmarking and Research in Multi-Objective Reinforcement Learning
Florian Felten, Lucas Nunes Alegre, Ann Nowe, Ana L. C. Bazzan, El Ghazali Talbi, Grégoire Danoy, and Bruno Castro da Silva, ‘A Toolkit for Reliable Benchmarking and Research in Multi-Objective Reinforcement Learning’, in Proceedings of the 37th Conference on Neural Information Processing Systems (NeurIPS 2023).
Hyperparameter Optimization for Multi-Objective Reinforcement Learning
F. Felten, D. Gareev, E.-G. Talbi, and G. Danoy, (2023). Hyperparameter Optimization for Multi-Objective Reinforcement Learning. arXiv, Oct. 25, 2023. doi: 10.48550/arXiv.2310.16487.
MO-Gym: A Library of Multi-Objective Reinforcement Learning Environments
L. N. Alegre, F. Felten, E.-G. Talbi, G. Danoy, A. Nowé, and A. L. C. Bazzan, “MO-Gym: A Library of Multi-Objective Reinforcement Learning Environments,” Proceedings of the 34th Benelux Conference on Artificial Intelligence BNAIC/Benelearn 2022
MORL/D: Multi-Objective Reinforcement Learning based on Decomposition
F. Felten, E.-G. Talbi, and G. Danoy, (2022). MORL/D: Multi-Objective Reinforcement Learning based on Decomposition. In Proceedings of International Conference on Optimization and Learning (OLA2022)
Metaheuristics-based Exploration Strategies for Multi-Objective Reinforcement Learning
Felten, F., Danoy, G., Talbi, E. and Bouvry, P. (2022). Metaheuristics-based Exploration Strategies for Multi-Objective Reinforcement Learning. In Proceedings of the 14th International Conference on Agents and Artificial Intelligence - Volume 2: ICAART, ISBN 978-989-758-547-0; ISSN 2184-433X, pages 662-673. DOI: 10.5220/0010989100003116.
Talks
It depends: Dealing with Multiple Objectives in (MA)RL
invited-talk at Université de Liège, Liège, Belgium
RTL: Drone research in Luxembourg
News article at My lab, Esch-sur-Alzette, Luxembourg
Hyperparameter Optimization for MORL
Conference talk at MODem workshop at ECAI, Krakow, Poland
Spotlight on young researchers
Vulgarization at My lab, Online
Presentation of Metaheuristics Based exploration strategies at ICAART
Conference talk at International Conference on Agents and Artificial Intelligence, Online