A0683
Title: CLT for generalized linear spectral statistics of large-dimensional sample covariance matrices and its application
Authors: Qing Yang - University of Science and Technology of China (China) [presenting]
Abstract: The generalized linear spectral statistics (GLSS) for high dimensional sample covariance matrices is introduced. The joint asymptotic normality of this statistic associated with various test functions is established when the dimension and the sample size are comparable under weak assumptions. As a natural application of the theory, a novel statistic is proposed based on GLSS to conduct hypothesis testing for spiked covariance matrices. Simulations indicate the accuracy and enhanced power of the proposed statistics and illustrate considerably better performance compared to the existing methods on various occasions.
