ABSTRACT
Many statistical analyses are done assuming that the underlying population is normal. When the dimension of the parameter space is at least two the multivariate normal distribution and the distributions that can be derived from it, e.g., the chi square or Wishart distributions, prove to be very useful. Frequently, a linear regression model is assumed to have normally distributed error terms. The least square estimators of the parameters then follow a multivariate normal distribution. A focal point of this book is different estimation procedures for the parameters of a multivariate normal distribution. To help make this book reasonably self contained and for ready reference the main properties of the multivariate normal distribution are summarized in Section 2.1.
