Premium
Data transformation in the evidential reasoning‐based decision making process
Author(s) -
Sonmez Mahmut
Publication year - 2007
Publication title -
international transactions in operational research
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.032
H-Index - 52
eISSN - 1475-3995
pISSN - 0969-6016
DOI - 10.1111/j.1475-3995.2007.00598.x
Subject(s) - evidential reasoning approach , multiple criteria decision analysis , ranking (information retrieval) , belief structure , computer science , process (computing) , transformation (genetics) , selection (genetic algorithm) , decision analysis , decision theory , management science , decision making , data mining , artificial intelligence , operations research , decision support system , mathematics , business decision mapping , engineering , statistics , biochemistry , chemistry , operations management , purchasing , gene , operating system
This paper describes the application of an evidential reasoning (ER)‐based decision making process to multiple‐criteria decision making (MCDM) problems having both quantitative and qualitative criteria. The ER approach is based on the decision theory and the theory of evidence and it uses the concept of ‘degree of belief’ to assess decision alternatives on each attribute. When faced with MCDM problems, evaluation and selection or ranking of alternatives appear to be both challenging and vital to arrive at a rational and robust decision. In the presence of both qualitative and quantitative evaluations in an MCDM problem, it is necessary, when using the ER‐based decision making process, to transform or convert quantitative data into a belief structure using a number of grades so that the converted belief structure and the original quantitative data are equivalent in values or utilities. This paper suggests three scenarios for data transformation and examines how the ranking of decision alternatives is changed when different scenarios of data transformation are used. Ranking of UK universities using the ER approach is illustrated as an example.