A1220
Title: Multivariate rank-based expectation of the conditional difference for testing independence
Authors: Xiaoli Kong - Wayne State University (United States) [presenting]
Abstract: The purpose is to introduce a class of multivariate rank-based measures for testing independence, which utilize the expected conditional characteristic function-based independence criterion. The tests are computationally efficient and remain well-defined under minimal assumptions on the underlying distributions. The asymptotic distribution for these test statistics is established, and their effectiveness is demonstrated through simulation studies. Additionally, the practical utility of the proposed test is illustrated through a real data application.
