A0613
Title: GLSS of sample correlation matrix
Authors: Xiao Han - University of Science and Technology of China (China) [presenting]
Abstract: A sample correlation matrix is an important random matrix used in high-dimensional data analysis. Motivated by this, spectral properties of large dimensional sample correlation matrices are studied by introducing generalized linear spectral statistics (GLSS). In particular, a joint CLT is established for this statistic related to various factors, which generalizes the result in a prior study. Meanwhile, the difference in GLSS is discovered between sample covariance matrices and sample correlation matrices. Extensive simulations verify the theoretical results.
