Sparse estimation of high-dimensional correlation matrices

Several attempts to estimate covariance matrices with sparsity constraints have been made. A convex optimization formulation for estimating correlation matrices as opposed to covariance matrices is proposed. An efficient accelerated proximal gradient algorithm is developed, and it is shown that this method gives a faster rate of convergence. An adaptive version of this approach is also discussed. Simulation results and an analysis of a cardiovascular microarray confirm its performance and usefulness.