Theory and Methods of Statistics

11

Linear Models

Abstract

The standard Gauss-Markov model is introduced, and estimation and inferential issues for this model are addressed. Model selection procedures such as Akaike Information Criterion, Bayesian Information Criterion, cross-validation, and generalized cross-validation are discussed. A brief introduction is given to alternative approaches for regression such as ridge, lasso, and partial least squares. Random- and mixed-effect models are introduced along with strategies for estimating the variance components and the nonrandom parts of such models. Many examples are given from regression, ANOVA, ANCOVA, and mixed linear models in order to clarify the key concepts discussed in this chapter.

Keywords

Gauss-Markov Models; Regression; ...