Discriminant function analysis; Linear discriminant function analysis
Discriminant analysis (DA) is one of a large class of methods for performing supervised classification. It uses a set of measured characteristics or attributes of subjects or objects to (a) put the observed units into one of two (or more) alternative groups a prioridefined or (b) formulate differing classes or groups. Thus, the method can be used with two rather distinct objectives. The first is as a predictive tool with the aim of formulating a rule that will permit objects to be classified into one of several predefined classes. The second is to help understanding (i. e., explanation or description) with the aim of building a model that helps us understand the structure in data. For example, is it possible to predict which patient's breast cancer will have spread to the surrounding lymph nodes? In what way do those infants at high risk of dying from Respiratory Distress Syndrome...