Research Interests

Our research lies at the intersection of causal inference, causal discovery, and causal decision-making, with the goal of building machine learning systems that are not only accurate but also understandable, fair, and actionable.

We develop methods that uncover and utilize causal structures in complex, non-stationary, and high-dimensional environments. Our core directions include causal effect identifiability and transportability — generalizing causal estimates across heterogeneous settings; causal decision-making — optimizing actions under uncertainty with causality as a first principle; causal machine learning — bridging causal representation learning with modern deep learning; and causal discovery — recovering causal graphs from observational, relational, and temporal data.

Beyond these, we are also interested in explainable and trustworthy AI — including fairness, robustness, and interpretability — and in emerging connections between causal reasoning and large-scale foundation models.

Research Projects

• Causal Machine Learning (NRF, PI, 2023 ~ 27 with Innovative Research Lab Initiation Grant)
• Self-Motivated AI: Developing self-directed AI agents that can solve new problems (IITP, Co-I, 2022 ~ 26, PI: Byoung-Tak Zhang)
• Center for Optimizing Hyperscale AI Models and Platforms (NRF, Co-I, 2023~, PI: Jaejin Lee)
• Metabolomic Big Data Analysis (MFDS, Co-I, 2023 ~ 25)
• Scalable Causal Discovery (LG AI Research, PI, 2025 ~ 26)
• Advancing Credit Decision Making (AFINIT, PI, 2025 ~ 26)

Past Projects

• Semantic Search for Korean Medical and Legal Documents (SNU, co-PI, 2022~23, PI: Hyopil Shin)
• Association and Causality in Metabolomic data (MFDS, co-PI, 2022)
• Supply Chain based Financial Keyword Analysis (NH Investment, 2021)
• Causal Discovery for Time Series (LG AI Research, PI, 2023.04~24.04)
• An algorithmic aspect of proxy-based causal inference (SNU, PI, 2021~24)
• Deep Generative Models for Causal Reasoning (LG AI Research, PI, 2024.05~25.05)