Research

My research focuses on developing novel Bayesian statistical methods and their theoretical properties, motivated by real biological applications. Methodologically, my interests revolve around graphical models, Bayesian nonparametrics and their applications to analyze high-throughput genetic data. In particular, I am strongly interested in covariate-dependent graphical models. Such model can facilitate the development of personalized therapeutic options to cancer patients based on the genetic markers, which has a profound impact on improving human life. Genetic data analysis has been the driving force behind most of my research projects. I am also interested in other Data Science problems, such as off-policy evaluation and meta-learning. I believe that a strong theoretical foundation lays the groundwork of any statistical methods. My theoretical work includes variable selection consistency and graph selection consistency for Bayesian hierarchical models. I enjoy discussing and collaborating with researchers from different fields such as geneticists, nutritionists and pharmaceutical scientists.