A Comparison of the Classification Accuracy of Linear and Quadratic Statistical Discriminant Models versus Linear and Quadratic Programming Discriminant Models

Author

Jerry William Ferry, University of Arkansas, FayettevilleFollow

Date of Graduation

8-1986

Document Type

Dissertation

Degree Name

Doctor of Philosophy in Business Administration (PhD)

Degree Level

Graduate

Department

Accounting

Advisor/Mentor

Thomas W. Jones

Committee Member

David E. Douglas

Second Committee Member

Nolan E. Williams

Keywords

Linear programming model, quadratic programming model, Monte Carlo techniques

Abstract

The purpose of this research was to compare the classification accuracy of two mathematical programming models versus traditional statistical discriminant analysis. Monte Carlo techniques were used to compute population 1, population 2, and average misclassification rates for the linear discriminant function (LDF), the quadratic discriminant function (QDF), a linear programming discriminant model (LPDM), and a quadratic programming discriminant model (QPDM) for specific values of several parameters which affect discriminant analysis. This study was restricted to the two group, two variable discriminant problem.

Citation

Ferry, J. W. (1986). A Comparison of the Classification Accuracy of Linear and Quadratic Statistical Discriminant Models versus Linear and Quadratic Programming Discriminant Models. Graduate Theses and Dissertations Retrieved from https://scholarworks.uark.edu/etd/3438