Title :
A new method for AR order determination of an ARMA process
Author :
Xiao, Chaung-Bai ; Zhang, Xian-Da ; Li, Yan-Da
Author_Institution :
Dept. of Autom., Tsinghua Univ., Beijing, China
fDate :
11/1/1996 12:00:00 AM
Abstract :
Using a combination of the minimum description length (MDL) criterion and overdetermined instrumental variable method, a computationally efficient method for AR order determination of an ARMA process is proposed. It is shown that the well-known singular value decomposition (SVD) method for AR order determination is a special case of the proposed method. Numerical simulations are presented to show the effectiveness of this method
Keywords :
Bayes methods; autoregressive moving average processes; information theory; numerical analysis; singular value decomposition; AR order determination; ARMA process; Bayesian information criterion; MDL criterion; SVD method; minimum description length; numerical simulations; overdetermined instrumental variable method; singular value decomposition; Adaptive signal processing; Analytical models; Eigenvalues and eigenfunctions; Instruments; Least squares approximation; Parameter estimation; Predictive models; Signal processing algorithms; Singular value decomposition; Speech;
Journal_Title :
Signal Processing, IEEE Transactions on
