Machine Learning & Intelligent Control Lab.

기계학습 및 지능형제어 연구실

관련기사 바로가기
기계학습 및 지능형제어 연구실

저희 연구실에서는 기존의 한계를 기계학습 알고리즘 개발, 수학적 지식을 기반으로 하는 이론적인 성능 개선, 기계학습의 실생활 적용에 집중하여 연구를 진행하고 있습니다. 저희가 AI 분야에서 주도하는 연구를 수행하기 위해 다음과 같은 연구 분야에 집중하고 있습니다.

Reinforcement Learning: 실생활에서 로봇이나 시스템을 최적으로 제어하기 위해, 환경에서 얻어지는 보상을 기반으로 최적의 제어 정책을 얻는 연구
특히, 강화학습 분야에서도, 실생활의 다양한 상황을 고려하기 위해 아래와 같은 세부 연구 분야들을 고려하여 연구를 진행하고 있습니다.
- Offline Reinforcement Learning
- Domain Adaptation/Imitation Learning
- Multi-Agent Reinforcement Learning
- Meta Reinforcement Learning
- Robust/Safe Learning
- Statistical Learning
- Intelligent Control Systems

Major research field

Reinforcement Learning, Offline RL, Domain Adaptation/Imitation Learning, Multi-Agent RL, Meta RL, Robust/Safe Learning, Intelligent Control Systems

Desired field of research

Real-world AI Application, General Machine Learning, Intelligent Control Systems

Research Keywords and Topics

- Offline Reinforcement Learning
Offline-RL learns policy by utilizing only the experience it gathers without additional interaction with the environment. There are still some challenges, such as distribution shifts and out-of-distribution problem.

- Domain Adaptation/Imitation Learning
Imitation learning is a branch of research aiming to apply reinforcement learning to real-life scenarios, focusing on learning policies that mimic the actions of experts. Recently, there has been progress in research on Domain Adaptation and Cross-Domain studies, enabling imitation of actions from experts in different domains,

- Multi-Agent Reinforcement Learning
In multi-agent RL, multiple agents aim to learn policies that would maximize the expected return from a shared environment. Coordination among the agents is essential for achieving this goal as the agents effect themselves as learning progresses.

Research Publications

* indicates corresponding author

- Proceedings of the AAAI conference on artificial intelligence (AAAI), “FoX: Formation-aware exploration in multi-agent reinforcement learning,” Yonghyeon Jo, Sunwoo Lee, Junghyuk Yum, Seungyul Han*, Vancouver, Canada, Feb. 2024.

- The 37th Conference on Neural Information Processing Systems (NeurIPS) 2023, “Domain Adaptive Imitation Learning with Visual Observtation,” Sungho Choi, Seungyul Han*, Woojun Kim, Jongseong Chae, Whiyoung Jung, Youngchul Sung, New Orleans, LA,USA, Dec. 2023.

- The 39th International Conference on Machine Learning (ICML), “Robust imitation learning against variations in environment dynamics,” Jongseong Chae, Seungyul Han*, Whiyoung Jung, Myungsik Cho, Sungho Choi, and Youngchul Sung, Jul. 2022.

Patents

- “시각적 관찰데이터를 활용한 도메인 적응형 모방학습 방법 및 시스템,” Youngchul Sung, Sungho Choi, Woojun Kim, Jongseong Chae, Whiyoung Jung, Seungyul Han, application number: 10-2023-0021570.

- “환경역학 변화에 강인한 모방학습 방법 및 시스템,” Youngchul Sung, Jongseong Chae, Whiyoung Jung, Myungsik Cho, Sungho Choi, Seungyul Han, application number: 10-2023-0021569.

국가과학기술표준분류

  • EE. 정보/통신
  • EE01. 정보이론
  • EE0108. 인공지능

국가기술지도분류

  • 정보-지식-지능화 사회 구현
  • 012300. 인공지능/지능로봇 기술

녹색기술분류

  • 녹색기술관련 과제 아님
  • 녹색기술관련 과제 아님
  • 999. 녹색기술 관련과제 아님

6T분류

  • IT 분야
  • 정보처리 시스템 및 S/W
  • 010314. 신호처리기술(영상/음성처리/인식/합성)