Roxana Rădulescu
Scholar

Roxana Rădulescu

Google Scholar ID: jue1nxAAAAAJ
Intelligent Systems, Utrecht University
Artificial IntelligenceMulti-agent systemsReinforcement LearningMulti-Objective Optimisation
Citations & Impact
All-time
Citations
1,351
 
H-index
13
 
i10-index
18
 
Publications
20
 
Co-authors
13
list available
Resume
Academic Achievements
  • Developed MOMAland, an open-source multi-objective multi-agent reinforcement learning Python library; will be giving a tutorial on multi-objective (multi-agent) decision-making at AAMAS 2024; published work on 'Emergent Cooperation under Uncertain Incentive Alignment'.
Research Experience
  • She has extensive experience in developing multi-agent decision systems, particularly where each agent is driven by different objectives and goals.
Education
  • Previously, she was a FWO Postdoctoral fellow at the Artificial Intelligence Lab, Vrije Universiteit Brussel, Belgium.
Background
  • Currently an Assistant Professor in AI and Data Science at the Department of Information and Computing Sciences, Utrecht University. Her research focuses on the development of multi-agent decision-making systems under the paradigm of multi-objective multi-agent reinforcement learning.
Miscellany
  • Offers several tutorials and lectures on multi-objective decision-making, including both single and multi-agent settings.