Selective Overview Discriminant Analysis

Abstract

In this paper we will review the concepts of discriminant analysis which is one of the important multivariate statistical analysis methods. It is used in various scientific, particularly the studies that are concerned with classification and prediction. It is having discriminant function which has an ability to distinguish between observations and then is classified the observations into two groups or more than it.

 It belongs to it depending on a special function called (the discriminant function), which is a linear combination between the independent variables and the dependent variable, which is of the nominal type. The main objective of this research was to address the subject of discriminant analysis from its early beginnings, present the studies and research that concerned it, how the linear discriminant function developed, and then propose new discriminatory functions such as the quadratic and logistic discriminant functions. It also presented the basic concepts of classification processes, conditions and assumptions that must be available to apply the discriminant analysis method, and then dealt with the Robust Discriminant Analysis method, which dealt with the problems of the presence of outliers among the data under study, which violate the basic assumptions of applying the discriminant analysis method, such as the hypothesis of the normal distribution of data.