The use of fuzzy outranking relations in preference modelling
Fuzzy Sets and Systems - Special issue dedicated to Professor Claude Ponsard
Generalizations of semiorders: a review note
Journal of Mathematical Psychology
A characterization of PQI interval orders
Discrete Applied Mathematics - Special issue: The 1998 conference on ordinal and symbolic data analysis (OSDA '98)
Expert Systems with Applications: An International Journal
Weapon selection using the AHP and TOPSIS methods under fuzzy environment
Expert Systems with Applications: An International Journal
Design a new mixed expert decision aiding system using fuzzy ELECTRE III method for vendor selection
Expert Systems with Applications: An International Journal
A fuzzy multicriteria methodology for selection among energy alternatives
Expert Systems with Applications: An International Journal
Hi-index | 12.05 |
So far, competing forecasting models are compared to each other using a single criterion at a time, which often leads to different rankings for different criteria - a situation where one cannot make an informed decision as to which model performs best overall; that is, taking all performance criteria into account. To overcome this methodological problem, we propose to use a Multi-Criteria Decision Analysis (MCDA) based framework and discuss how one might adapt it to address the problem of relative performance evaluation of competing forecasting models. Three outranking methods have been used in our empirical experiments to rank order competing forecasting models of crude oil prices; namely, ELECTRE III, PROMETHEE I, and PROMETHEE II. Our empirical results reveal that the multidimensional framework provides a valuable tool to apprehend the true nature of the relative performance of competing forecasting models. In addition, as far as the evaluation of the relative performance of the forecasting models considered in this study is concerned, the rankings of the best and the worst performing models do not seem to be sensitive to the choice of importance weights or outranking methods, which suggest that the ranks of these models are robust.