Publications
I have published papers in AI and ML conferences (e.g., UAI, NeurIPS, ICML, AAAI, AISTATS). Within the scope of causal inference, I investigated the problem of causal effect identifiability allowing causal inference engines to take more diverse sets of data. In parallel, I also studied on the identifiability problem for heterogeneous domains (similar to the underlying ideas of domain adaptation or transfer learning in ML) called transportability . I also pursue research in decision-making where causality serves as the first principle to solve the problem. During my PhD study, I mostly spent time on understanding causal discovery from relational data such as relational conditional independence, relational Markov equivalence classes, relational causal discovery algorithms both theoretical and practical. Recently, I am interested in the intersection of causality and machine learning (e.g., causal representation learning) and causal discovery.
Working Papers
- LLM-Guided Causal Discovery (Juhyeon+)
- Causality-inspired Domain Generalization (Dong Kyu)
- Sequential Adjustment Criterion (YJ)
- Causality in Rested Bandit (ND, Yeahoon, Soungmin)
- Robust Differences-in-Differences (Jeong Ha)
- Value of Information under Insolubility (RC, RE, EB, Minwoo)
- Citation for Retrieval Augmented Generation (Juhyeon+, submitted)
- Causal Inference under Weaker Assumption (Yesong, Yeahoon, Inwoo, submitted)
- Fine-grained Causal Dynamics Learning (Inwoo, submitted)
- Robust Monte-Carlo Tree Search (Inwoo+, submitted)
- Representation Learning for Instrumental Variables (Jung Soo+, submitted)
- Regularized Synthetic Control (Jeongsup+, submitted)
- Robust Causal Discovery (Jonghwan+, submitted)
Published Papers
* for joint first authorship
Filter, Rank, and Prune: Learning Linear Cyclic Gaussian Graphical Models
Soheun Yi, Sanghack Lee
AISTATS 2024 [accepted]
Learning to ignore: Single Source Domain Generalization via Oracle Regularization
Dong Kyu Cho, Sanghack Lee
Causal Representation Learning Workshop at NeurIPS 2023 [paper]
Quantized Local Independence Discovery for Fine-Grained Causal Dynamics Learning in Reinforcement Learning
Inwoo Hwang, Yunhyeok Kwak, Suhyung Choi, Byoung-Tak Zhang, Sanghack Lee
GenPlan 2023: Seventh Workshop on Generalization in Planning at NeurIPS 2023 [paper]
Causal Dynamics Learning with Quantized Local Independence Discovery
Inwoo Hwang, Yunhyeok Kwak, Suhyung Choi, Byoung-Tak Zhang, Sanghack Lee
The Second Workshop on Spurious Correlations, Invariance and Stability at ICML 2023 [paper]
On Discovery of Local Independence over Continuous Variables via Neural Contextual Decomposition
Inwoo Hwang, Yunhyeok Kwak, Yeon-Ji Song, Byoung-Tak Zhang, Sanghack Lee
The 2nd Conference on Causal Learning and Reasoning (CLeaR) 2023 [paper]
Detecting Causality by Data Augmentation via Part-of-Speech tagging
Juhyeon Kim, Yesong Choe and Sanghack Lee
CASE Workshop at EMNLP 2022
Counterfactual Transportability: A Formal Approach
Juan D. Correa, Sanghack Lee and Elias Bareinboim
ICML 2022 [paper]
Partition-based Local Independence Discovery
Inwoo Hwang, Byoung-Tak Zhang, and Sanghack Lee
Causal Inference Challenges in Sequential Decision Making: Bridging Theory and Practice Worksop at NeurIPS 2021
Nested Counterfactual Identification from Arbitrary Surrogate Experiments [paper]
Juan D. Correa, Sanghack Lee and Elias Bareinboim
NeurIPS 2021 [paper]
Causal Identification with Matrix Equations [paper]
Sanghack Lee and Elias Bareinboim
NeurIPS 2021
Characterizing Optimal Mixed Policies: Where to Intervene and What to Observe [paper], [slides], [poster]
Sanghack Lee and Elias Bareinboim
NeurIPS 2020
Causal Effect Identifiability under Partial-Observability [paper]
Sanghack Lee, and Elias Bareinboim
ICML 2020
General Transportability — Synthesizing Experiments from Heterogeneous Domains [paper]
Sanghack Lee, Juan D. Correa, and Elias Bareinboim
AAAI 2020
Identifiability from a Combination of Observations and Experiments [paper], [slides]
Sanghack Lee, Juan D. Correa, and Elias Bareinboim
AAAI 2020
General Identifiability with Arbitrary Surrogate Experiments [paper] [errata]
Sanghack Lee, Juan D. Correa, and Elias Bareinboim
UAI 2019, Best Paper Award
Towards Robust Relational Causal Discovery [paper]
Sanghack Lee and Vasant Honavar
UAI 2019
Fairness in Algorithmic Decision Making: An Excursion Through the Lens of Causality [paper]
Aria Khademi, Sanghack Lee, David Foley, and Vasant Honavar
WWW 2019
On Structural Causal Bandit with Non-manipulable Variables [paper], [poster], [slides]
Sanghack Lee and Elias Bareinboim
AAAI 2019
Structural Causal Bandits: Where to Intervene? [paper], [code], [poster]
Sanghack Lee and Elias Bareinboim
NeurIPS 2018
Pre-Ph.D. —
A Kernel Conditional Independence Test for Relational Data [code], [paper]
Sanghack Lee and Vasant Honavar
UAI 2017
Self-Discrepancy Conditional Independence Test [code], [paper]
Sanghack Lee and Vasant Honavar
UAI 2017
A Characterization of Markov Equivalence Classes for Relational Causal Model with Path Semantics [code], [paper], [appendix]
Sanghack Lee and Vasant Honavar
UAI 2016
On Learning Causal Models from Relational Data [code] [paper]
Sanghack Lee and Vasant Honavar
AAAI 2016
“Teens are from Mars, Adults are from Venus”: Analyzing and Predicting Age Groups with Behavioral Characteristics in Instagram [paper]
Kyungsik Han, Sanghack Lee, Jin Yea Jang, Yong Jung, and Dongwon Lee
WebSci 2016
Lifted Representation of Relational Causal Models Revisited: Implications for Reasoning and Structure Learning [paper]
Sanghack Lee and Vasant Honavar
UAI 2015 Workshop on Advances in Causal Inference
Transportability from Multiple Environments with Limited Experiments [paper]
Elias Bareinboim*, Sanghack Lee*, Vasant Honavar, and Judea Pearl
NeurIPS 2013
Causal Transportability of Experiments on Controllable Subsets of Variables: z-Transportability [paper]
Sanghack Lee and Vasant Honavar
UAI 2013
m-Transportability: Transportability of a Causal Effect from Multiple Environments
Sanghack Lee and Vasant Honavar
AAAI 2013
Learning Classifiers from Distributional Data
Harris Lin*, Sanghack Lee*, Ngot Bui*, and Vasant Honavar
IEEE Second International Congress on Big Data