Multivariate classification models

In the previous articles, we have gone on a journey, starting from basic statistical concepts such as p values and Type 1 and 2 errors to confusion matrices and classification. Examples have been univariate, in that we have modelled classes using a single variable, with a univariate cut-off as to the decision criterion according to whether a sample either from a test set or as an independent prediction, is a member of predefined classes or groups.