A MIXED INTEGER PROGRAMMING APPROACH FOR DATA CLASSIFICATION BASED ON CASE-BASED DISTANCE METHOD

 

Li-Ching Ma

 

Abstract

 

Classification is a procedure to separate data or alternatives into two or more than two classes. The need to classify alternatives involving multiple criteria into one of distinct classes is considerable in practice. Therefore, determining how to assist decision makers in classification is an important issue of multi-criteria decision aids. This study tries to develop a classification model incorporating the advantages of the cases-based distance method and mixed integer programming model. The proposed classification model can classify alternatives by evaluating sets of cases selected by decision makers, reduce misclassification rate, and improve impact of outliers. The proposed classification approach is demonstrated using an inventory planning and control example.

 

Lecture Notes in Management Science (2011) Vol. 3: 47-56

3rd International Conference on Applied Operational Research, Proceedings

© Tadbir Operational Research Group Ltd. All rights reserved.

www.tadbir.ca

 

ISSN 2008-0050 (Print)

ISSN 1927-0097 (Online)

 

ARTICLE OUTLINE

 

·         Introduction

·         The Conventional Case-Based Distance Method For Classification

·         The Proposed Approach

·         A Numerical Example

·         Conclusions  

·         References

 

Full Text PDF