|
Jordan algebras and statistical inferences for parameters
in mixed linear and nonlinear models The lecture will present some main results about estimation and testing hypotheses for parameters in mixed models. The main conclusion is that we can unified theory about estimation and testing hypotheses both for fixed parameters and variance components. The statistical inference will be reduced to the minimal sufficient and complete statistics. For nonlinear estimation we present a new numerical procedure using full information about Hessian for getting LSE.
Qualquer informação acerca dos seminários pode ser obtida por
correio electrónico para seminarios@mat.ubi.pt.
Departamento de Matemática da Universidade da Beira Interior |