Kazuaki Yamada

The University of Tokyo, Toyo University, Kobe University

Papers

7

Total Citations

28

H-Index

4

About

Kazuaki Yamada is a robotics and artificial intelligence researcher whose work centers on reinforcement learning, autonomous robot behavior acquisition, and multi-robot cooperative systems. Over more than two decades of research, Yamada has made consistent contributions to solving one of robotics' fundamental challenges: enabling robots to learn effective sensor-to-motor mappings through trial-and-error rather than explicit programming. His early work on Instance-Based Classifier Generators (1999) laid groundwork for reactive behavior learning, while subsequent investigations demonstrated how neural networks and CMAC architectures could replace memory-intensive lookup tables in Q-learning frameworks, making reinforcement learning more practical for real-world, high-dimensional environments. Yamada's research extends naturally into multi-robot coordination, exploring how distributed reinforcement learning units can enable cooperative manipulation and scalable swarm behaviors. More recently, his interest has broadened to human-robot interaction, including an insightful study analyzing pedestrian gaze behavior to inform the design of socially intuitive mobile robots. His publications, cited across the robotics and machine learning communities, reflect a career dedicated to bridging theoretical learning algorithms and practical autonomous systems — valuable reading for students interested in intelligent robotics and adaptive control.

Research Focus

Key Achievements

4
H-Index
7
Papers
28
Total Citations
4
Avg Citations/Paper
🏆 Most Cited Paper
Network Parameter Setting for Reinforcement Learning Approaches Using Neural Networks
6 citations · 2011
📈 Most Prolific Year: 2011 (1 Papers)
🤝 Key Collaborators: 10
🏛 Institutions: The University of Tokyo, Toyo University, Kobe University

Top Papers

  1. 1
  2. 2
  3. 3
  4. 4
  5. 5
  6. 6
  7. 7

Key Collaborators

Contact & Links

Available for collaboration
Content generated · 15 days ago