BIT Makes New Progress in Optimal Block Design Theory

  Recently, assistant professor Kong Xiangshun from School of Mathematics and Statistics, Beijing Institute of Technology and his collaborators published a research paper entitled "Approximate and Exact Designs for Total Effects" on Annals of Statistics, one of the four top international journals of Statistics. This paper presents a consistent optimal design method for several traditional models in block design. 

  In drug trials, taking drugs on the same day not only has an influence on the treatment effect on the day, but also has an impact on the treatment effect on the next day. Therefore, when researchers give each patient in clinical trial of the first stage (the block in the design theory) a different drug and a placebo day by day to estimate the pesticide effect, they need to take into account the effects of the drug taken the day before. The of statistical test design problems of similar structure also exist in many fields such as agricultural seed selection and industrial processing method selection.

  The above paper by assistant professor Kong Xiangshun and his coauthors, whcih is based on optimal design theory, gives a scientific conclusion on how to construct a consistent optimal design scheme under various criteria for a given number of blocks. When there are t schemes, n blocks and p design points in each block, the number of possible designs is as high as tnp. By using the new optimal design theory proposed by assistant professor Kong Xiangshun and his collaborators, the consistent optimal design scheme under various criteria can be found among tnp possibilities. This method not only enriches the theory of optimal block design in experimental design, but also provides scientific and economic schemes for relevant medical, agricultural and industrial experiments.

  The research was done by assistant professor Kong Xiangshun in collaboration with associate professor Zheng Wei of The University of Tennessee and his students, and was funded by National Natural Science Foundation of China and the Academic Launch Program for Young Teachers of BIT.
Link of the paper: https: //imstat. org/journals-and-publications/annals-of-statistics/annals-of-statistics-future-papers/
Profile of the research team and individual:

  The experimental design team of School of Mathematics and Statistics, Beijing institute of technology has actively carried out cooperative research and academic exchanges both at home and abroad, and the team members have established a long-term relationship of exchange and cooperation with famous domestic and foreign experimental design scholars, such as academician C.F.Jeff Wu, professor Roshan Vengazhiyil Joseph, professor Ai Mingyao, associate professor Zheng Wei and etc, showing a strong momentum of development.

  Kong Xiangshun, assistant professor,  key member of  the experimental design team from School of Mathematics and Statistics, BIT.  He graduated from Nankai University with a bachelor's degree and Peking University with a doctor's degree, was a visiting scholar at The University of Tennessee. He mainly engaged in experimental design and related statistical research work. In addition, he has published several high-level academic papers in prestigious statistical journals such as Annals of Statistics, Technometrics, Statistica Sinica and Computational Statistics & Data Analysis.


​Release Date: 2020-08-23News Source: School of Mathematics and Statistics
Editor: News Agency of BIT