The linear sufficiency in multivariate Gauss-Markov model

Rok, strany: 1996, 143 - 147

The notion of linearly sufficient statistics (LS) has been introduced into the literature by Baksalary and Kala (1981) [J. K. Baksalary, R. Kala: Linear transformations preserving best linear unbiased estimators in a general Gauss–Markov model, Ann. Statist. 9 (1981), 913–916]. They defined it as such transformation of the fixed, onevariate Gauss-Markov model which preserves information needed for the estimation in this model. The aim of the present paper is generalization of the notion LS from one onto multivariate case.

ISO 690:

Kornacki, A. 1996. The linear sufficiency in multivariate Gauss-Markov model. In Tatra Mountains Mathematical Publications, vol. 7, no.1, pp. 143-147. 1210-3195.

APA:

Kornacki, A. (1996). The linear sufficiency in multivariate Gauss-Markov model. Tatra Mountains Mathematical Publications, 7(1), 143-147. 1210-3195.