Research Area

Reinforcement Learning

Machine learning is a basic tool for making decisions or predictions based on incomplete data. In this area, SISReL is currently focusing on reinforcement learning, which will be a major tool for AI robots, smart cities and autonomous vehicle, from various research perspectives such as 

• improved value estimation

• enhancing exploration

• intrinsic reward design for sparse-reward reinforcement learning 

• meta and multi-task reinforcement learning

• domain adaptation

• parellel learning

• imitation learning

• multi-agent reinforcement learning

• multi-objective reinforcement learning

• partially-observable Markov decision processes (POMDP).

Next Generation Communication Systems: 6G and Internet-of-Things

In this area, SISReL is conducting research on wireless communication systems and networks and their fusion with smart machine intelligence systems like connected vehicle from the perspective of real applications such as 6G and internet-of-things with extensive real world experience of the advisor. We are trying to come up with new algorithms, multi-access methods or system architectures with significant performance improvement for wireless communication networks. Currently, we are investigating new antenna technology for 6G, flexible network topology, AI-aided communications for future evolution of wireless technologies.