본문 바로가기 대메뉴 바로가기

한밭대학교컴퓨터공학과

HIGHHANBAT

RLLab

RLLab

About the Lab

Reinforcement learning, an optimization technique for solving sequential decision-making problems, is garnering attention for achieving general-purpose AI. It is being utilized in various fields such as robot control, autonomous driving, gaming, and natural language processing, with innovations achieved in each field. Our lab aims to contribute to the advancement of AI technology through continuous research and development and industrial application of reinforcement learning and data analysis technologies. We are also endeavoring to continuously expand the scope of our research through interdisciplinary research.

Advisor: Soo-Young Jang

Thesis

  1. S. Jang and H. Kim, “Efficient deep reinforcement learning under task variations via knowledge transfer for drone control,” ICT Express, 2024.
  2. S. Jang and H. Kim, “Supervised pre-training for improved stability in deep reinforcement learning,” ICT Express, 2023.
  3. S. Jang and J. Lee, “Utilizing hidden observations to enhance the performance of the trained agent,” IEEE Robotics and Automation Letters, 2022.
  4. S. Jang and C. Choi, “Prioritized environment configuration for drone control with deep reinforcement learning,” Human-centric Computing and Information Sciences, 2022.
  5. 5. S. Jang and H. Kim, “Entropy-aware model initialization for effective exploration in deep reinforcement learning,” Sensors, 2022.

Patents

  1. Method and Apparatus for Controlling the Movement of Real Objects Using an Intelligent Agent Trained in a Virtual Environment (Registered, 10-2355021, Electronics and Telecommunications Research Institute)
  2. Communication Device and Method for Optimizing TCP Congestion Window (Registered, 10-2389104, Electronics and Telecommunications Research Institute)
  3. Apparatus and Method for Controlling Data Transmission in Communication Systems (Registered, 10-2359274, Samsung Electronics)
  4. Connection Control Device and Method in Communication Systems (Registered, 10-2333144, Samsung Electronics)
  5. Communication method and apparatus for optimizing TCP congestion window (Registered, US 11178056)