Recently published a paper on Multi-Objective Deep Reinforcement Learning with Lexicographic Task-priority constraints, advocating for less reward engineering in favor of constrained RL.
Research Experience
Appointed as leader of the WASP Reinforcement Learning cluster from 01.03.23, which is a group of ~50 RL and Robotics researchers.
Education
Currently a Ph.D. student at the Adaptive and Interpretable Learning Systems Lab, funded by the WASP program, at Örebro University.
Background
Research interests: Deep Reinforcement Learning, Transfer Learning, and Robotics. Believes that the DRL framework has immense potential for industry automation and optimization, but the intransparency and data inefficiency of deep neural network-based AI systems must be addressed for safe and efficient real-world employment of these technologies.
Miscellany
Interests include contributing to the community and sharing insights about topics he cares about, especially where there is a lack of material or documentation.