Abstract：An ideal independence test should possess three properties: it should be zero-independence equivalent, numerically efficient, and asymptotically normal. We introduce a  slicing procedure for estimating a popular measure  of nonlinear dependence, leading to the resultant sliced independence test  simultaneously possessing all three  properties.   In addition, the power performance of the sliced independence test improves as  the number of observations within each slice increases. The popular rank test corresponds to a special case of the sliced independence test that contains two observations within each slice. The sliced independence test is thus more powerful than the rank test. The size performance of the sliced independence test is insensitive to the number of slices, in that the slicing estimation is consistent  and asymptotically normal for a  wide range of  slice numbers.     We further adapt the sliced independence test to account for the presence of multivariate control variables.   The  theoretical properties are  confirmed using comprehensive simulations and an application  to an astronomical data set.

报告人简介：

朱利平，中国人民大学杰出学者特聘教授、博士生导师，统计与大数据研究院院长，国家杰出青年科学基金获得者，国家重大人才工程计划入选者。长期从事复杂高维、超高维和非线性相依数据分析基础理论、方法和应用研究工作。
现任中国现场统计学会高维数据分会和生存分析分会副理事长。先后担任统计学领域国际顶级学术期刊《Annals of Statistics》、国际重要学术期刊《Statistica Sinica》和《Journal of Multivariate Analysis》等国际学术期刊Associate Editor，以及《系统科学与数学》和《应用概率统计》等国内重要学术期刊编委、副主编。