Peter Stone
AT&T (United States), The University of Texas at Austin, Carnegie Mellon University, Laboratoire d'Informatique de Paris-Nord, Sony Computer Science Laboratories, Sony Corporation (United States), Sony (Taiwan), FZI Research Center for Information Technology, Li Creative Technologies (United States)
Papers
259
Total Citations
10,570
H-Index
51
About
Peter Stone is a pioneering researcher at the intersection of machine learning, multiagent systems, and robotics, whose work has fundamentally shaped how autonomous agents learn, cooperate, and compete in complex environments. His landmark 2000 survey on multiagent systems from a machine learning perspective, now boasting over 1,100 citations, established a foundational framework that continues to guide researchers across both fields. Stone's concept of "Layered Learning," introduced through his highly influential book and accompanying papers, demonstrated how hierarchical machine learning structures can enable sophisticated teamwork in real-time, adversarial settings — insights drawn partly from robotic soccer competitions. His policy gradient reinforcement learning approaches to quadrupedal locomotion, cited nearly 600 times, showcased how robots could physically optimize their own movement through experience rather than hand-engineering. Stone further advanced the field with his formulation of "ad hoc autonomous agent teams," addressing the critical challenge of collaboration among agents with no prior coordination. His work on cross-domain transfer learning and mobile robot navigation surveys underscores a career-long commitment to making autonomous systems adaptable, efficient, and practically deployable across diverse real-world scenarios.
Research Focus
Key Achievements
Top Papers
- 1Multiagent Systems: A Survey from a Machine Learning Perspective1,188 citations · 2000
- 2Policy gradient reinforcement learning for fast quadrupedal locomotion585 citations · 2004
- 3
- 4Layered Learning in Multiagent Systems381 citations · 2000
- 5Ad Hoc Autonomous Agent Teams: Collaboration without Pre-Coordination326 citations · 2010
- 6Layered Learning in Multiagent Systems: A Winning Approach to Robotic Soccer263 citations · 2000
- 7
- 8Cross-domain transfer for reinforcement learning204 citations · 2007
- 9Machine learning for fast quadrupedal locomotion151 citations · 2004
- 10Layered Learning144 citations · 2000