Title :
Moving average processes and maximum entropy
Author :
Politis, Dimitris Nicolas
Author_Institution :
Dept. of Stat., Purdue Univ., West Lafayette, IN, USA
fDate :
5/1/1992 12:00:00 AM
Abstract :
A characterization of the stochastic process that has maximum entropy among all moving average processes of order q, subject to the condition that the autocovariances γ(k) satisfy γ(k)=ck, for k=0,1,. . ., p, is provided by exploiting properties of the inverse autocovariance sequence
Keywords :
correlation theory; entropy; information theory; parameter estimation; spectral analysis; stochastic processes; autocorrelation; autocovariances; inverse autocovariance sequence; maximum entropy; moving average processes; spectral estimation; stochastic process; Autocorrelation; Decoding; Encoding; Entropy; Gaussian processes; Laplace equations; Lattices; Speech; Stochastic processes; Vector quantization;
Journal_Title :
Information Theory, IEEE Transactions on
