Discrete Filtered Fractals in Signal Modeling and ProcessingUniversity of Minnesota, 1992 - 286 pages |
Table des matières
PRELIMINARIES | 11 |
MAXIMUM LIKELIHOOD ESTIMATION OF | 61 |
FILTERED fdGn PROCESSES IN SIGNAL MODELING | 83 |
Droits d'auteur | |
4 autres sections non affichées
Expressions et termes fréquents
ak+1 Akaike Information Criterion approximate MLE ARMA models ARMA processes autocorrelation function autocorrelation matrix BH(t compute converge correlation function covariance method CR bound Cramer-Rao bound differenced Gaussian noise EM algorithm estimate of parameter estimate the parameters Expectation-Maximization algorithm Expected Loglikelihood expected value Figure filtered fdGn models filtered fractals fractal dimension fractional Brownian motion fractionally differenced Gaussian Gaussian noise process Hence increments inversion iterative algorithm large data sets Levinson algorithm likelihood function linear prediction coefficients log-likelihood function long term correlation Mandelbrot maximization Maximum Likelihood Estimation mean square error ML estimate noise variance nonlinear observation set parameter estimation parameter H polynomial prediction error power probability density procedure proposed rational filter reflection coefficients sample path second derivative self-similar shaping filter short term correlation signal processing spectral density spectrum stationary stationary process technique Temperature Variations term correlation structures vector wa(n white noise Yule-Walker equations