|Taku MORIYAMA||Last modified date：2019.01.18|
Assistant Professor / Institute of Mathematics for Industry
Field of Specialization
Statistical science,Nonparametric inference
I study the asymptotic theory of nonparametric statistical analysis. Nonparametric methods are robust to model assumptions, and can handle data from unspecified distributions. Assuring the accuracy needs large enough sample size since nonparametric methods only use information from realizations. Therefore, making inference about the tail behavior of a distribution, where the data has little information, is a difficult problem in general. Moreover, the accuracy drastically deteriorates in high dimensional cases, when the curse of dimensionality occurs and obtained realizations get dispersed. I aim to devise and implement nonparametric methods with practical accuracy, and try improvement of statistics from various perspectives.