Discrimination Methods For Solving Recovery and NonRecovery Oil Problems

Abstract

Discriminant analysis is one of the most classical classification procedure used to discriminate and differentiate groups as possible by using one or more attributes. It also assigns observations to one of the pre-defined and separated groups. There are many parametric discriminant statistical methods such Fisher’s FDA, quadratic QDA, and logistic LoDA methods, and non-parametric operation researches methods such as maximize minimum of deviations MMD, and minimize sums of deviation MSD methods to discriminate the difference between groups and classify new observations to the suitable group. Operation researches methods are fixable and do not require any assumptions as statistical methods. This study introduces a comparison between three statistical methods and two linear programming models for solving recover and nonrecover oil problems. Section (1) includes an introduction of the study. Section (2) provides a suitable review of statistical and operation researches methods. Section (3) introduces an application of some discrimination methods for solving recovery and nonrecovery Oil data problems. Section (4) includes the analysis and recommendation of the study. Finally, at the end of the study, there is the references section in section (5).