My research focuses on the development and application of modern statistical methods for biomedical and public health research. I am particularly interested in survival analysis, longitudinal data analysis, causal inference, missing data, and statistical learning, with applications to chronic disease epidemiology and clinical prediction.
My research integrates methodological biostatistics with collaborative biomedical research. I develop and apply statistical methods for analyzing complex health data arising from observational studies, longitudinal cohorts, and survival studies. My current work combines classical statistical methodology with modern machine learning techniques to improve disease risk prediction and clinical decision-making.
This research program uses data from the National Health and Nutrition Examination Survey (NHANES) 1999–2018 to investigate mortality risk among U.S. adults through the development and application of modern statistical and machine learning methods. The program combines traditional survival analysis with statistical learning techniques to improve prediction, understand disease progression, and support evidence-based public health decision-making.
Investigating the association between glycated hemoglobin (HbA1c) and all-cause mortality among U.S. adults using NHANES. This research examines nonlinear relationships, time-varying effects, and subgroup differences using survival analysis and flexible regression methods to improve understanding of diabetes- related mortality risk.
This research program uses data from the Health and Retirement Study (HRS) to investigate the development of chronic disease among older U.S. adults. The work focuses on longitudinal patterns of health, repeated measurements, and risk factors associated with incident diabetes.
My methodological research focuses on the development of statistical methods for survival analysis, missing data, and complex sampling designs. This work emphasizes semiparametric inference, efficient estimation, and robust statistical methodology motivated by real-world biomedical and public health applications.
My doctoral dissertation developed semiparametric additive hazards models for survival data with missing covariates under two-phase sampling designs, motivated by HIV vaccine efficacy research. The work introduces augmented inverse probability weighted estimation procedures with strong theoretical properties and practical applications to biomedical studies.