We aim to train researchers who can build data-driven optimization systems using statistical and machine learning techniques, and collaborate effectively with domain experts across diverse fields. Our group works closely with Robot Intelligence Lab, co-led with Sungjoon Choi, enabling interdisciplinary research spanning statistical AI, reinforcement learning, and robotics. Through this collaboration, students learn to bridge theory and practice, and develop systems that operate in real-world environments. Since the lab is jointly operated, prospective students may contact either kyungjae_lee at korea dot ac dot kr or sungjoon-choi at korea dot ac dot kr. We welcome students who are interested in intelligent decision-making under uncertainty and real-world impact.

I am currently an associate professor in the Department of Statistics at Korea University, Seoul. Prior to this, I served as an assistant professor in the Department of Artificial Intelligence at Chung-Ang University (2021–2024). Before then, I earned my B.S. in Electrical Engineering and Computer Science from Seoul National University in 2015, followed by a Ph.D. in Electrical and Computer Engineering from the same university in 2021 under the supervision of Prof. Songhwai Oh.

• Black-Box Optimization 

  • Bandit Algorithms: Heavy-tailed bandits, Minimax optimal heavy-tailed bandits

  • Bayesian Optimization: Heavy-tailed BO

  • Combinatorial Optimization: Perturbation-based exploration for MCTS, Learning-based boosting

• Reinforcement Learning and Its Variants

  • Regularized Markov Decision Processes: Maximum Tsallis entropy RL (and its special case)

  • Preference-Based Reinforcement Learning (RL from Human Feedback): Sequential comparison for efficient PbRL

  • Safe Reinforcement Learning: Gradient aggregation for multiple constraints, Spectral risk safe RL

  • Distributional Reinforcement Learning: Pertrubed exploration for dist. RL, Eluder dimension-based regret analysis for dist. RL

• Imitation Learning

  • Inverse Reinforcement Learning: IRL with negative demonstrations

  • Generative Adversarial Imitation Learning: Maximum Tsallis entropy framework

• Smart Assembler: Robot Active Learning for Unseen Parts Assembly (2023 ~ 2028)

  • Funded by NRF [우수신진연구, 최초혁신실험실]

• Towards An Industrial World Model: Development of Self-Evolving Surrogate Learning for Process Simulation (2026~2027)

  • Funded by LG AI research

• Multi-Suction Cup Grasping Network (2025~2026)

  • Funded by CMES Robotics

• Development of Emotion Recognition/Generation-based Interacting Edge Device Technology for Mental Health Care

  • Funded by IITP

• Development of Stock Prediction Models using Explainable AI with Multi-modal Data and Deep Reinforcement Learning

  • Funded by NRF

• Development of a lightweight gripper and an object recognition system

  • Funded by Seoul Business Agency

• Development of AI Algorithm for Operator Assistance

  • Funded by Doo-San

• Development of Reinforcement Learning Algorithm for Urban Autonomous Driving

  • Funded by Hyundai NGV

• CAN Data-based Forklift CUP Analysis

  • Funded by Doo-San

• Uncertainty Aware Task and Motion Planning

  • Funded by NRF

• Reinforcement Learning based Press Molding Optimization

  • In collaboration with Hyundai NGV

• Domain Generalization with Pre-trained Models

  • In collaboration with Kakao Brain

  • This work was published at ECCV 2022

• Developing Embedded AI Systems

  • Funded by Doo-San

• Domain Generalization via Finding Flat Minima

  • In collaboration with Naver Clova AI 

  • This work was published at NeurIPS 2021

• Tsallis Entropy Reinforcement Learning

  • In collaboration with Kakao Brain 

  • This work was published at RSS 2020

Department of Statistics (Korea University)

• Ph.D.

  • Taeheon Lim (Federated Learning, Machine Unlearning)

• M.S.

  • Jungwoo Hong

  • Injoon Song

Department of AI (Korea University)

Co-led with Prof. Sungjun Choi

• Students affiliated with the lab can be found on the lab website: Robot Intelligence Lab

Department of AI (Chung-Ang University)

• Integrated Ph.D.

  • Hyeonjun Park (Multi-armed bandit, Reinforcement learning, Heavy-tail estimation)

• Ph.D.

  • No Ph.D. graduates yet

• M.S.

  • Taeheon Lim (Now a Ph.D. student at Korea University)

  • Minseok Bae (Now at Khameleon)

  • Jonghyeon Kim (Now at LG AI Research Internship)

  • Junho Kim (Now at KETI)

  • Youngjoon Kwon ()

  • Juhan Park (Now a Ph.D. student at Korea University)

  • Youngtaek Hong (Now at Wonik Robotics)

  • Chanwoo Kim (Now a Ph.D. student at Korea University)

  • Sangho Yoon (Now at Guardian AI)

  • Jaewon Kim (Now at SFA)

  • Daejong Jin (Now at HD Hyundai Robotics)

  • Hyunjun Yang (Now at Rebellions)

  • Jaesung Ryu (Now at Guardian AI)

  • Joonhyeok Lee (Now at hutom)