Research

My research lies at the intersection of statistics, machine learning, and public health. I develop methods for causal inference from incomplete, biased, high-dimensional, and otherwise imperfect data, with the goal of improving the reliability of scientific evidence and supporting informed decision making. For a publication list, visit Google Scholar.

Research themes Grant support Research code

Research themes

The themes below separate methodological questions by the type of imperfection or structure in the data.

01

(Un)measured confounding

Graphical identification, semiparametric estimation, and sensitivity analysis when important causes are latent or unobserved.

 02

Missing data

Identification and estimation of causal and statisical parameters when observations are coarsened, or selectively observed and missing not at random.

 03

Constrained learning, algorithmic fairness

Statistical learning and decision-making methods that satisfy fairness, policy, or other causal and statistical constraints.

 04

High-dimensional data

Causal dimension reduction methods for multi/high-dimensional confounders, exposures, or mediators.

 05

Mediation analysis

Methods for inferring mediation and path-specific effects in complex longitudinal and multi-mediator settings.

 06

Health and public health data science

Applications of causal and statistical methods to public health, medicine, and biomedical science.
(Un)measured confounding

Principled inference under (un)measured confounding

Selected publications

  • [Paper] Controllable Generative Sandbox for Causal Inference. arXiv, 2026.
  • [Paper] Coarsening Bias from Variable Discretization in Causal Functionals. UAI, 2026.
  • [Paper] Causal Inference with the "Napkin Graph". arXiv, 2025.
  • [Paper] Average Causal Effect Estimation in DAGs with Hidden Variables: Beyond Back-Door and Front-Door Criteria. arXiv, 2024.
  • [Paper] Semiparametric Sensitivity Analysis: Unmeasured Confounding in Observational Studies. Biometrics, 2024.
  • [Paper] Flexible Nonparametric Inference for Causal Effects under the Front-Door Model. arXiv, 2023.
  • [Paper] Causal Inference in the Presence of Interference in Sponsored Search Advertising. Frontiers in Big Data, 2022.
  • [Paper] On Testability of the Front-Door Model via Verma Constraints. UAI, 2022.
  • [Paper] Semiparametric Inference for Causal Effects in Graphical Models with Hidden Variables. JMLR, 2022.
  • [Paper] Causal Inference Methods For Bias Correction In Data Analyses. PhD Thesis, 2021.
Missing and cencored data

Principled inference under informative missingness

Selected publications

  • [Paper] Weighting-Based Identification and Estimation in Graphical Models of Missing Data. arXiv, 2026.
  • [Paper] Self-separated and Self-connected Models for Mediator and Outcome Missingness in Mediation Analysis. Statistical Science, 2026.
  • [Paper] MissNODAG: Differentiable Learning of Cyclic Causal Graphs from Incomplete Data. TMLR, 2026.
  • [Paper] Response to Discussions of "Causal and Counterfactual Views of Missing Data Models" Statistica Sinica, 2026.
  • [Paper] Causal and Counterfactual Views of Missing Data Models. Statistica Sinica, 2026.
  • [Paper] Graphical Models of Entangled Missingness. arXiv, 2023.
  • [Paper] Sufficient Identification Conditions and Semiparametric Estimation under Missing Not at Random Mechanisms. UAI, 2023.
  • [Paper] On Testability and Goodness of Fit Tests in Missing Data Models. UAI, 2023.
  • [Paper] Full Law Identification in Graphical Models of Missing Data: Completeness Results. ICML, 2020.
  • [Paper] Identification In Missing Data Models Represented By Directed Acyclic Graphs. UAI, 2020.
Constrained learning, algorithmic fairness

Decision making under constraints such as fairness norms

Selected publications

  • [Paper] Bridging Prediction and Intervention Problems in Social Systems. arXiv, 2025.
  • [Paper] Fair Risk Minimization under Causal Path-Specific Effect Constraints. arXiv, 2024.
  • [Paper] Statistical Learning for Constrained Functional Parameters in Infinite-Dimensional Models. arXiv, 2024.
  • [Paper] Optimal Training of Fair Predictive Models. CLeaR, 2022.
  • [Paper] Learning Optimal Fair Policies. ICML, 2019.
  • [Paper] Fair inference on outcomes. AAAI, 2018.
High-dimensional confounders, exposures, mediators

Causal dimension reduction in highdimensional settings

Selected publications

  • [Paper] Causal Sufficient Dimension Reduction for Multiple Continuous Exposures with an Application to Environmental Mixtures. arXiv, 2026.
  • [Paper] Semiparametric Causal Sufficient Dimension Reduction of Multidimensional Treatments. UAI, 2022.
Mediation and pathway analysis

Understanding the mechanisms through which effects operate

Selected publications

  • [Paper] Self-separated and Self-connected Models for Mediator and Outcome Missingness in Mediation Analysis. Statistical Science, 2026.
  • [Paper] Assessing Racial Disparities in Healthcare Expenditures via Mediator Distribution Shifts. Statistics in Medicine, 2026.
  • [Paper] Estimation of Personalized Effects Associated With Causal Pathways. UAI, 2018.
Methods and applications in public health data science

Causal methods for complex public health questions

Collaborative causal and statistical research in public health, medicine, and biomedical science, spanning vaccine effectiveness, environmental exposures, health disparities, neuroimaging, and clinical decision making.

Vaccine efficacy and clinical epidemiology

  • [Paper] Causal Vaccine Effects on Post-infection Outcomes in the Naturally Infected. arXiv, 2026.
  • [Paper] Target trial emulation without matching: a more efficient approach for evaluating vaccine effectiveness using observational data. arXiv, 2025.
  • [Paper] Vaccine efficacy against naturally asymptomatic infections: A novel estimand for quantifying vaccine effects. arXiv, 2025.

Air pollution and enviromental epidemiology

  • [Paper] Causal Sufficient Dimension Reduction for Multiple Continuous Exposures with an Application to Environmental Mixtures. arXiv, 2026.

Health disparities and equity

  • [Paper] Assessing Racial Disparities in Healthcare Expenditures via Mediator Distribution Shifts. Statistics in Medicine, 2026.

Neuroimaging and preprocessed outcomes

  • [Paper] Causal Inference for Preprocessed Outcomes with an Application to Functional Connectivity. arXiv, 2026.
  • [Paper] Deep dag learning of effective brain connectivity for FMRI analysis. IEEE International Symposium on Biomedical Imaging, 2023.
  • [Paper] Learning task-aware effective brain connectivity for FMRI analysis with graph neural networks. IEEE International Conference on Big Data, 2022.

Grant Support

Selected support for methodological research in causal inference, missing data, statistical learning, and public health data science.

NSF · MPI · 2025–2028
Two Sides of a Tapestry: Causal Inference and Model Discovery Amid Information Gaps in Complex Data
NIH/NIBIB T32EB035514 · MPI · 2025–2030
Advancement of Diagnostics for a Just Society Training Program (ADJUST)
NIH/NIEHS R21ES036795 · PI · 2024–2026
Causally-Sufficient Dimensionality Reduction Methods for Assessing Joint Effects of Air Pollution Mixtures on Health Outcomes
Georgia CTSA BERD Pilot · PI · 2022–2024
Tackling Missing Data from a Causal and Counterfactual Point of View
Selected Collaborative Support . CO-I
Collaborative grants in diagnostics, infectious disease modeling, real-world clinical decision making, and functional connectivity estimation. Projects include: CDC CAMP 3 and CIDMATH, PCORI personalized lung cancer therapy with real-world data, and NIH/NIMH functional connectivity methods.

Research code

For code associated with a specific paper, please reach out to us.