An Evaluation of Covariance and Correlation Analysis in Entropy Method

Abstract

Objectives:The main aim of this research is to obtain the criterion weight in Multi-Criteria Decision Making (MCDM) with conflicting criteria, as the relevance and impact of the criterion weight alters the outcome of any decision-making. Methods: This study proposes a modification to the existing Entropy technique using principal component analysis, which is found to be suitable for all dayto-day real-life problems. Findings: A comparative study is done between covariance and correlation analysis to validate the accuracy of the proposed technique. The proposed modification is illustrated with a numerical example. Novelty: The inclusion of the concept of covariance analysis in the principal component analysis is a new approach to the determination of the criterion weight in MCDM.