Ph.D. in Statistics

Ph.D. in Statistics

Research Associate in Biostatistics at the University of Pennsylvania

Huiyuan Wang

Research Associate in Biostatistics · University of Pennsylvania

I develop statistical methods for reliable learning and inference when data are biased, imperfect, distributed, or structurally complex.

My research spans causal inference and real-world evidence, distributed inference, statistical foundations for adaptive and verifiable AI, and learning from structured data. I am also interested in active learning, retrieval-augmented systems, benchmarking and evaluation of large language models, continual learning, and learning systems designed to avoid shortcut solutions.

News

Selected updates
  1. Upcoming MOSAiC at JSM 2026 Invited Paper Session 1094

    Join us on Wednesday, Aug 5, 2026, 2:00–3:50 PM, in Room CC-253C at the Boston Convention & Exhibition Center.

  2. Talks ICSA 2026 presentation in Shenzhen

    I presented Efficient Inference for Distributed Data with Structural Missingness at the ICSA Applied Statistics Symposium in Shenzhen, China.

  3. Accepted MOSAiC accepted as a JASA Discussion Paper

    Delighted to share that MOSAiC was accepted as a Discussion Paper in JASA's Applications and Case Studies section.

  4. Published Robust watermark detection published in JRSS-B

    Our work on watermark detection under human edits is now published in Volume 88, Issue 2, pages 491–515.

  5. Recognition Top Reviewer at NeurIPS 2025

    Honored to be recognized among the top 10% of reviewers.

  6. Talks Three presentations at JSM 2025

    Glad to have presented work on calibrated digital twins, augmented target trial emulation, and negative-control-calibrated difference-in-differences in Nashville.

  7. Published Negative-control-calibrated difference-in-differences published in npj Digital Medicine

    Our work on calibration for time-varying unmeasured confounding in real-world data is now published.