• Research
  • Statistical Methods and Calculation in Big Data and Design of Experoments


    Big data is revolutionizing our life, work and thinking. An important and imperative research project is how to mine and use the information in big data efficiently. Our research about big data and design of experiments will focus on the following four sub-projects: 1) big data mining and algorithm; 2) the statistical inference of high dimensional convariance matrix; 3) the monitoring and diagnostics for the on-line big data; 4) the design theory and modelling for complex experiments. The detailed research contents will include: 1) in the view of computer science: data fusion algorithms, data mining and recommendation algorithms, privacy preserving data mining algorithms and parallel algorithms; 2) in the view of statistics: the robust estimation of high dimensional and sparse covariance matrix, the robust two-sample t test for high dimensional data, on-line monitoring and diagnostics for high dimensional data stream, how to define and monitor the quality of big data, the on-line monitoring for multivariate functional data, the optimality theory and construction of supersaturated designs with high and mixed levels, the construction and modelling of various Latin hypercube designs, etc. Our research results will include patents or software copyrights, papers and the cultivation of PhD and Master students.