2022

chapter in monograph

english

EN We review the Multiple Criteria Decision Aiding (MCDA) methods in the stream of Robust Ordinal Regression (ROR). They incorporate indirect preference information in the form of decision examples and verify the consequences of applying all compatible instances of an assumed preference model. We focus on four aspects distinguishing the ROR approaches: considered problem typologies, forms of accepted preference information, employed decision models, and types of delivered outcomes quantifying the robustness of results. We also discuss significant extensions of ROR. The most prevailing ones include Stochastic Ordinal Regression, active learning strategies, algorithms for generating dedicated explanations of the decision outcomes, and procedures for answering questions regarding the stability of results. Finally, we list selected real-world applications of ROR in various fields. The review confirms the position of ROR as one of the essential methodological streams in MCDA in the last decades.

09.02.2022

185 - 205

10.1007/978-3-030-96318-7_10

https://link.springer.com/chapter/10.1007/978-3-030-96318-7_10

Intelligent Decision Support Systems : Combining Operations Research and Artificial Intelligence - Essays in Honor of Roman Słowiński

20