Atsushi Miyamae
Papers
2
Total Citations
16
H-Index
2
About
Atsushi Miyamae is a researcher whose work lies at the intersection of reinforcement learning, evolutionary computation, and robotics control. His primary research focuses on developing advanced policy search methods for complex, nonholonomic systems—robotic platforms whose motion constraints make them notoriously difficult to control through traditional theoretical derivation. Miyamae’s major contribution is the introduction of instance-based policy learning, a novel approach that leverages real-coded genetic algorithms to directly optimize control policies. This method bypasses the need for precise mathematical models, offering a practical solution for real-world robotic control problems where theoretical guarantees are often unattainable. His most cited work, "Instance-based Policy Learning by Real-coded Genetic Algorithms and Its Application to Control of Nonholonomic Systems" (2009, 13 citations), demonstrates the effectiveness of this approach, while his earlier paper (2008, 3 citations) formalizes the optimization framework. Though his citation counts are modest, Miyamae’s work is notable for bridging the gap between direct policy search in reinforcement learning and the practical demands of nonholonomic robot control, providing a foundation for data-driven, model-free control strategies in challenging environments.
Research Focus
Key Achievements
Top Papers
- 1
- 2