ABSTRACT
Linear models (Chapter 2) have been extended to GLMs (Chapter 2) and linear mixed models (Chapter 5). Two extensions are combined in HGLMs (Chapter 6). GLMs can be further extended to joint GLMs, allowing structured dispersion models (Chapter 3). This means that a further extension of HGLMs can be made to allow structured-dispersion models (Chapter 7) and to include models for temporal and spatial correlations (Chapter 8). With DHGLMs it is possible to allow heavy-tailed distributions in various components of HGLMs. This allows many models to be unified and further extended, among others those for smoothing (Chapter 9), frailty models (Chapter 10) and financial models. All these models are useful for the analysis of data.
