Tinghua Chen

I’m currently a Ph.D. candidate at Penn State University, advised by Justin Silverman (Committee Chair), and Cory McCartan. My research develops reliable machine learning and causal inference methods for imperfect data, with applications in biology, health, and the social sciences.

During summer 2026, I interned at Google as a Research Data Scientist, working with Joey Dickens and Fei Li. Before pursuing my Ph.D., I worked as a statistician at the FDA in the Center for Drug Evaluation and Research (CDER). I also earned a concurrent master’s degree in Applied Statistics and International Affair from Penn State.

Causal Machine Learning for Imperfect Data: I develop robust, scalable, and theoretically grounded methods for causal inference when data are incomplete, noisy, or aggregated. My work focuses on reliable models for settings with latent variables, measurement error, and unobserved confounding.

Scalable Bayesian Modeling: I build interpretable models for complex longitudinal count and compositional data, including generalized Gaussian processes and generalized dynamic linear models. This work pairs flexible modeling with efficient inference through a parallelized C++ backend and R interface.

Scientific and Social Applications: I apply these methods to data-rich scientific settings where the underlying signals are only partially observed, including microbiome and genomic sequence count data, as well as social science problems involving aggregate and observational data, such as elections.

Under Review

1. Bayesian Multinomial Logistic Normal Additive Regression for Microbiome Data Analysis (2025). Under journal revision. With Michelle Pistner Nixon, Lawrence A. David, and Justin D. Silverman.

Published Papers

2. Scalable Inference for Bayesian Multinomial Logistic-Normal Dynamic Linear Models (2025). AISTATS. With Manan Saxena and Justin D. Silverman.
3. Watermarking Counterfactual Explanations (2025). PRICAI, best paper award. With Hangzhi Guo, Firdaus Ahmed Choudhury, and Amulya Yadav.
4. Generalizing the “Masterpiece Effect” in fine art pricing: Quantile Hedonic regression results for the South African fine art market, 2009–2021 (2023). Economic Modelling, 124, 106301. With Johannes W. Fedderke.

View all publications on Google Scholar →