Our research direction：
(1) Factorial experiments
Orthogonal design of factorial experiments has been widely used in biological test or industrial product development process. We mainly study the construction theory and algorithm of orthogonal design. On the other hand, we also study factorial experiments which are non-orthogonal design but high efficiency. In addition, we provide a new statistical method to test the dispersion effect that affects the variability of response and research the optimality of the factorial experiments to estimate dispersion effect. Recently, we focus on optimality and construction of replicated factorial experiments design.
(2) Interval estimation for conformance proportions
Interval estimation for conformance proportions is usually applied in environmental control, bioequivalence, quality control in industrial production processes, and even animal and plant breeding. However, in the past, interval estimation for conformance proportions was limited to simple distributions. In our research, we aimed to extend the construction method to a broader distribution. The research results mostly matched the practical application.
(3) Genomic selection
he first step of genomic selection is building genomic selection model (GS model) for continuous, ordinal, categorical data of training population. Second, we use the GS model to predict the genomic estimated breeding value (GEBV) of each individual of test population (GEBV, genomic estimated breeding value). Finally, we conduct selection according to GEBV. It involves different large-p-small-n data analysis. Recently, our lab propose the new methods to search important SNP markers and decide train population.