Title :
Estimation of partial correlation matrices using Cholesky decomposition
Author :
Dickinson, Bradley W.
Author_Institution :
Princeton University, Princeton, NJ, USA
fDate :
4/1/1979 12:00:00 AM
Abstract :
A method for obtaining estimates of the partial correlation matrices of a multivariate process using Cholesky decomposition is described. The estimates provide a minimum phase autoregressive model for the observed data and a positive definite autocorrelation function estimate.
Keywords :
Autoregressive processes; Correlation methods; Least-squares estimation; Spectral estimation; Calculus; Covariance matrix; Filtering; Kernel; Matrix decomposition; Polynomials; Smoothing methods; Stochastic processes; Stochastic systems; Technological innovation;
Journal_Title :
Automatic Control, IEEE Transactions on
DOI :
10.1109/TAC.1979.1102008
