Data Depth based Discriminant Classification Analysis

Abstract

The Data Depth is used to measure the depth or area of any variation according to the distribution base. This results in the average natural centre-outer of the sample points. The essence of the deep procedure in multivariate analysis is to measure the degree of centrality of points associated with assumptions or probability distributions. This working data examines in-depth methods for determining the size of the site, ie. deepest or focal point. In addition, various in-depth procedures are studied in real and simulation contexts using R software. The performance of various data-depth processes is analyzed with numerical description by calculating the average misclassification error as part of a discriminative analysis.