Seunghan Lee이승한

I am an AI Scientist at LG AI Research. I received my integrated M.S./Ph.D. in Statistics and Data Science at Yonsei University, co-advised by Prof. Taeyoung Park (DSLab) and Prof. Kibok Lee (ML Lab). Previously, I was a Deep Learning Engineer / Researcher at KRAFTON and a research intern at Naver Cloud (HyperCLOVA X) and SK Telecom.

My research centers on deep learning for time series — including multimodal time-series forecasting, time-series foundation models, financial time series, and anomaly detection. I am also interested in representation / self-supervised learning and diffusion models.

I write about AI / Data Science on my blog (1,400+ posts, 2020.03–2026.01).

Current Work

• EXAONE Finance: TS foundation model for financial assets
• EXAONE Demand: TS foundation model for demand forecasting
• Financial TS forecasting (Equity, Macro, Fundamental, etc ..)
• Multimodal TS forecasting (TS+Text)
• Supply chain knowledge graph
• TS reasoning model
• Explainable TS forecasting

Experience

Education

Research Interests

• Deep learning for time series
  • Multimodal time series forecasting [W6, C7, C8/W8]
  • Time-series foundation models [W4, C5, W7, T2, T3, W11, W12]
  • Time-series reasoning, explainability & agents [P1, C8/W8, W10]
  • Representation learning [C1/W1, C2/W2]
  • Others (e.g., SSL, Diffusion, Anomaly Detection) [C3, C4, W3, W5, P3]
• AI for finance
  • Financial time series [C7, C8/W8, T1, T2, W14, P2]
  • Financial language models [C9, W13]
• Deep learning for tabular / graphs [C6, W9, J2]

Publications

28 papers  ·  20 as first author

CConference9 (7) WWorkshop14 (10) JJournal2 (1) TTech. report3 (2) PPreprint3 (3) Total (first-author)

* Equal contribution   † Co-corresponding author