Satinder Singh
Papers
1
Total Citations
30
H-Index
1
About
Satinder Singh is a pioneer in reinforcement learning and artificial intelligence, whose work has fundamentally shaped how agents learn and make decisions in complex, multi-agent environments. His research spans core areas of machine learning, including reinforcement learning, multi-agent systems, and the development of algorithms that enable autonomous agents to operate effectively under uncertainty. Singh is best known for his foundational contributions to hierarchical reinforcement learning and the study of agent architectures, particularly through his influential work on the "3 vs. 2 Keepaway" soccer domain, which provided a rigorous testbed for multi-agent coordination and learning. This 2001 paper, with over 30 citations, demonstrated how reinforcement learning could be applied to real-time, continuous-state problems, inspiring a generation of researchers in robotics and game AI. Beyond this, his broader body of work—accumulating tens of thousands of citations—includes seminal papers on temporal-difference learning, bounded rationality, and the theory of reinforcement learning. A professor at the University of Michigan and a former director of research at Google DeepMind, Singh has also been recognized with multiple best paper awards and is a Fellow of the AAAI. His research continues to drive advances in autonomous systems, from game-playing agents to real-world decision-making tools.
Research Focus
Key Achievements
Top Papers
- 1Reinforcement Learning for 3 vs. 2 Keepaway30 citations · 2001