Akito Sakurai
Papers
1
Total Citations
2
H-Index
1
About
Akito Sakurai is a researcher whose work lies at the intersection of reinforcement learning and sensor selection, with a particular focus on enabling autonomous systems to operate efficiently under uncertainty. His most-cited paper, "Online Sensor Selection in Reinforcement Learning Environment" (2005), addresses a fundamental challenge in robotics: that more sensors do not always lead to better state representations. Sakurai proposed a novel multi-armed bandit formulation to allow a mobile robot to dynamically select an appropriate subset of sensors while simultaneously learning a state-action function. This contribution is pivotal for reducing computational overhead and improving learning efficiency in real-world environments. Though his citation count is modest—with the paper garnering 2 citations—the work is notable for its early recognition of the sensor selection problem within reinforcement learning, a topic that has since gained significant traction. Sakurai's research is particularly valuable for students and engineers working on resource-constrained autonomous systems, as it bridges the gap between theoretical bandit algorithms and practical robotic control.
Research Focus
Key Achievements
Top Papers
- 1Online Sensor Selection in Reinforcement Learning Environment2 citations · 2005