A0537
Title: Large dynamic covariance matrix modeling via adaptive local/global thresholdings
Authors: Shaojun Guo - Renmin University of China (China) [presenting]
Abstract: The main aim is to develop a general class of sparse dynamic covariance matrix models for capturing the dynamic information in large covariance matrices. We apply adaptive local/global thresholding techniques to recover sparsity structures. We consider a bias-corrected local linear smoother to estimate the large covariance matrix locally, which is shown to be very useful for threshold selection in adaptive/global thresholdings. The nonasymptotic concentration bounds of the resulting estimators under different functional sparsity scenarios are established. We also demonstrate that our proposed method significantly outperforms possible competitors through intensive simulation studies and is also applied to a real data set, revealing some interesting findings.
