A1372
Title: On distance correlation based independence testing with high-dimensional incomplete data
Authors: Jelena Radojevic - University of Belgrade (Serbia) [presenting]
Bojana Milosevic - University of Belgrade (Serbia)
Abstract: A distance correlation-based methodology for testing independence between high-dimensional random vectors with incomplete observations is presented. Under the assumption that data are missing completely at random (MCAR), the asymptotic null distributions of the proposed test statistics are established. The statistics are applicable in high-dimensional settings where the dimension is comparable to or exceeds the sample size. Finite-sample performance is investigated through extensive simulation studies under a variety of dependence structures, including nonlinear alternatives. Additionally, the behavior of the proposed methods under different missingness mechanisms is examined.
