论文部分内容阅读
In this talk block circular matrices which can be used to model dependence structure of hierarchical or multilevel data are considered. Several spectral properties of circular symmetric Toeplitz matrices will be outlined which are useful for inference about these models. Special focus is on matrices having Kronecker structure since they arise in many applications for modelling spatio-temporal data. Furthermore, estiMability of variance-covariance components of such matrices will be discussed.