Abstract:
We prove a theorem which makes it possible to construct likelihood equations for the estimation of parameters determining the covariance matrix of one-dimensional Gaussian sequences. Relationships are derived for the Fisher information matrix and the probability density function of the estimators.