Summary of Research
Early in my academic career, I worked on times series analysis. In the past few years, my primary research interest is in spatial statistics (which includes the spatio-temporal case), and the applications of spatial statistics to environmental, agricultural and natural resources sciences. My research has been supported by the National Science Foundation (DMS-0405782
, 2004-2007, DMS-0706835
, 2007-2010) and other sources. The research problems in spatial statistics that I have studied are roughly grouped as follows.
- Spatial Generalized Linear Mixed Models
- Infill Asymptotics
- Multivariate Spatial Statistics
- Methods for analyzing massive spatial data
Some Ongoing Projects
Approximate inferences for massive spatial or spatio-temporal data: The sample size of a spatial sample may be extremely large and this is usually the case with spatio-temporal data. Currently, I am studying some approximate approaches that are computationally feasible and yet yields no or little loss of efficiency compared with that of the MLE.
Bias correction for the approximate inferences for spatio-temporal data: Recent years have seen the development of new space-time covariance functions. Estimation of parameters in the covariance function can be a computational burden especially when the sample size is extremely large. I am working on an approach to employ misspecified but simpler space-time covariance functions and then adjust for the bias resulted from the misspecification. The adjustment is based on asymptotic results.
Stationary processes with skewed marginal distributions: In many applications, spatial data are skewed (and often right skewed as in many environmental studies). Sometimes directly modeling the skewed process is preferable to the transformation method.
I have collaborated with several agricultural scientists on various projects ranging from crop insurance, precision agriculture, plant pathology and natural resources. Many of the projects provided motivating spatial data.
- My collaborators at Purdue University are from several departments and centers: Forestry and Natural Resources, Agronomy, Agricultural and Biological Engineering, Earth and Atmospheric Sciences, Center for Environment and Purdue Climate Change Research Center.
I have more than 10 years of experience in statistical consulting, providing statistical advices on experimental design, data analysis, model selection and diagnosis, and various statistical issues.