A comparative study between the linear discriminant analysis method and the core discriminant analysis method - an applied study

Abstract

In this research, two methods were used to classify data and they are Linear Discriminant Analysis and method Kernel Discriminant Analysis, and they are the two most statistical methods used in data analysisas requiring these methods provide a number of hypothesesThe most important of these is that the explanatory variables are distributed in a multivariate normal distribution. This research aims to find the discriminatory function of each,And used as a function of classification (discrimination) of the two methods among patients,Where the use of real data for the two groups of patients infected and non-infected and The results were obtained using the Statistical Program (SPSS) for the linear discriminant analysis methodand (R – Package) for the kernel discriminant analysis method. andIt was found that the method of linear discriminant analysis is better because it gives less error classification of data. as it has been a comparison between the two modes in accordance with the standard classification error probability of (Misclassification).