Distributed Reinforcement Learning for Smart Gird Energy System
This is a project which applies reinforcement learning methods for energy management. All data I use are from real-time historical log data from the open energy system in the hillside housing area at OIST campus site.
Emergence of communication among multi reinforcement learning agents
This research project use a decentralized reinforcement learning framework to understand how communication emerges in learning agents and the meaning evolved along the learning process.
This is a project which I built robot platform with smartphones. The robots could be used for a variety of research topics, such as benchmarks for data-efficient reinforcement learning, and learning to cooperate with symbolic communication.