|
Seoungbin Bae I am a third-year Ph.D. student in the Department of Industrial & Systems Engineering at KAIST, advised by Prof. Dabeen Lee. Before that, I received my M.S. in Electrical Engineering from KAIST, where I worked on computational imaging under the supervision of Prof. Young-Gyu Yoon. My research lies in machine learning for sequential decision-making under uncertainty. I develop online learning, bandit, and reinforcement learning algorithms, with an emphasis on provable guarantees. My recent work includes nonlinear contextual bandits, constrained reinforcement learning, and bandit algorithms for queueing systems. CV / Google Scholar / Email / LinkedIn |
Publications(*: equal contribution) |
|
Algorithm for Contextual Queueing Bandits with Rate-Optimal Queue Length Regret
Seoungbin Bae, Dabeen Lee arXiv, 2026 |
|
Chebyshev Center-Based Direction Selection for Multi-Objective Optimization and Training PINNs
Hoyeol Yoon, Seoungbin Bae, Nam Ho-Nguyen, Dabeen Lee arXiv, 2026 |
|
Neural Logistic Bandits
Seoungbin Bae, Dabeen Lee ICML, 2026 |
|
Logistic Bandits with \(\tilde{O}(\sqrt{dT})\) Regret Without Context Diversity Assumptions
Seoungbin Bae, Dabeen Lee arXiv, 2026 |
|
Near-Optimal Primal-Dual Algorithm for Learning Linear Mixture CMDPs with Adversarial Rewards
Kihyun Yu, Seoungbin Bae, Dabeen Lee arXiv, 2026 |
|
Abunmix Enables the Simple and Robust Multiplexed Immunofluorescence Imaging
Woonggi La*, Seoungbin Bae*, Junyoung Seo*, Hayeong Yu*, Junmo Cho, Hyunwoo Kim, Hoyeon Nam, Seungjae Han, Euiin Yi, Eunsu Kim, Chan Kang, Hyejin Shin, Chang Woo Song, Young-Gyu Yoon, Jae-Byum Chang VIEW, 2026 |
|
Learning to Route and Schedule LLMs from User Retrials via Contextual Queueing Bandits
Seoungbin Bae, Junyoung Son, Dabeen Lee arXiv, 2026 |
|
Primal-Dual Policy Optimization for Linear CMDPs with Adversarial Losses
Kihyun Yu, Seoungbin Bae, Dabeen Lee ICLR, 2026 |
|
Queue Length Regret Bounds for Contextual Queueing Bandits
Seoungbin Bae, Garyeong Kang, Dabeen Lee ICLR, 2026 |
|
Highly Accurate Image Registration for 3D Multiplexed Cyclic Imaging Using Dense Labeling in Expandable Tissue Gels
Hyunwoo Kim, Joon-Goon Kim, Jueun Sim, Hoyeon Nam, In Cho, Hyejin Shin, Junyoung Kwon, Dae-Hyeon Song, Seoungbin Bae, Young-Gyu Yoon, Taeyun Ku, Jae-Byum Chang PLOS Biology, 2025 |
|
IMPASTO: Multiplexed Cyclic Imaging Without Signal Removal via Self-Supervised Neural Unmixing
Hyunwoo Kim*, Seoungbin Bae*, Junmo Cho, Hoyeon Nam, Junyoung Seo, Seungjae Han, Euiin Yi, Eunsu Kim, Young-Gyu Yoon, Jae-Byum Chang bioRxiv, 2022 |
|
Secure Key Exchange Method via Ill-Conditioned Inverse Matrix in Wireless Local Area Networks
Seoungbin Bae, Minwoo Joo, Wonjun Lee KSC, 2020 |
|
Poster Abstract: Suppressing CSI Leakage in Multi-user MIMO Networks via Precoding
Seoungbin Bae, Youngki Kim, Heejun Roh, Wonjun Lee IEEE INFOCOM Workshops, 2020 |
|
Multiplexed Immunofluorescence Imaging via Signal Unmixing
Patent applications [patent1] [patent2] |
Experience |
|
Research Intern, Department of Cyber Defense, Korea University
2019-2021
|
Teaching Experience |
|
Advanced Optimization for Data Science (DS801), Spring 2024
Matrix Computations for Signal Processing (EE548), Spring 2023 Introduction to Electronics Design Lab (EE305), Fall 2022, 2023 Wireless Mobile Communication Networks, Fall 2020 |
Academic Service |
| Reviewer, NeurIPS 2026 |
Awards and Honors |
|
The National Scholarship for Science and Engineering
Korea Student Aid Foundation (KOSAF) Full tuition during undergraduate years |
|
Military Scholarship
Ministry of National Defense, Republic of Korea Full tuition during undergraduate years |
| Template from here |