Area of Interest
My research in this area is to study different sampling design (RDD, cellphone, response dependent sampling and multi-frame and complex sampling), help collect data from sampled respondent, and develop innovative statistical methods to handle sampling issue and obtain accurate estimates. Survey sampling is a commonly used approach for collecting data in cancer genetics and epidemiology, especially when interest focuses on studying cancer-related problems in some particular population, even hard-to-reach subjects.
My methodological research primarily focuses on semiparametric models and high-dimensional data arising from a variety of statistical or biostatistical areas, including cancer survival or relapse time data analysis, longitudinal biomarker analysis, clinical trials, high-dimensional genetic data, and medical diagnostics. In the past ten years, I have published extensively in leading statistical journals.
My work has been internationally recognized and I have been invited to present his research all over the world. I have served as co-investigator on multiple NIH grants. Particularly, my research contribution to semiparametric models and inference mainly lies in extensive work on developing transformation models for the analysis of various types of censored data which largely exist in cancer-related events (death and relapse). Particularly, I proposed general transformation models for modeling counting processes, clustered survival data, recurrent events, cured survival, and joint analysis of recurrent events and terminal event.
My other research area lies in developing correct and efficient inference for emerging applications. Particularly, he and his co-author(s) developed an efficient likelihood approach for analyzing haplotype-environment interactions with uncensored or censored outcomes and under cohort or case-control designs in a discussion paper. We further considered haplotypes analysis in nested case-control or case-cohort study. We also developed a correct approach for handling second phenotype in a biased sample design and showed the equivalence between meta-analysis and mega-analysis in genome-wide association. Some of my recent work also studies reinforcement learning in non-small cell lung cancer trial and disease surveillance. In addition to methodology development, my research also includes developing sample size/power calculation for case-cohort studies with rare and non-rare disease, developing efficient computing algorithms for transformation models, and developing efficient sampling approaches for non-smooth estimation.
Awards and Honors
- 2011 Elected ASA Fellow
- 2010 Elected IMS Fellow
- 2008 Roy Kuebler Fund Award, Department of Biostatistics, The University of North Carolina
- 2008 Noether Young Scholar Award, American Statistical Association
- 2006 Junior Faculty Development Award, The University of North Carolina
- 2002 Center for AIDS Research Development Award, The University of North Carolina
- 2001 Travel Award, Rackham Graduate School, The University of Michigan
- 1997 Qualify Exam Excellence Award, Dept. of Statistics, The University of Michigan
- 1993 Guo-Moruo Award, The University of Science and Technology of China