Weighting Data using the Robust Transformation Matrix to Eliminate the Effect of Outliers in a Robust Discriminant Analysis Model

Abstract

Robust statistical methods are of great importance in statistical studies because they provide great resistance in the presence of basic violations of statistical analysis models due to failure to achieve one of the basic assumptions such as the normal distribution of data and others. One of the most important problems facing the researcher is the problem of the presence of outliers in the data under study. Therefore, the goal of this study is to reduce the impact of outliers on the accuracy of the results. The discriminant analysis method was applied to a set of data taken from the Iraqi Stock Exchange, where the outliers were weighted with certain weights to eliminate their impact on the results. The banks under study were classified into two groups based on the cut-off point. The classification error of the mentioned methods was measured and the results were good and reliable.