Pregled bibliografske jedinice broj: 939201
Empirical Comparison of three Methods for Approximating DEX Utility Functions
Empirical Comparison of three Methods for Approximating DEX Utility Functions // Proceedings of the 13th International Symposium on Operational Research SOR15 / Stirn, L. Zadnik ; Žerovnik J. ; Borštnar Kljajić M ; Drobne S. (ur.).
Ljubljana: Slovensko društvo Informatika. Sekcija za operacijske raziskave, 2015. str. 29-34 (predavanje, međunarodna recenzija, cjeloviti rad (in extenso), znanstveni)
CROSBI ID: 939201 Za ispravke kontaktirajte CROSBI podršku putem web obrasca
Naslov
Empirical Comparison of three Methods for
Approximating DEX Utility Functions
Autori
Mihelčić, Matej ; Bohanec, Marko
Vrsta, podvrsta i kategorija rada
Radovi u zbornicima skupova, cjeloviti rad (in extenso), znanstveni
Izvornik
Proceedings of the 13th International Symposium on Operational Research SOR15
/ Stirn, L. Zadnik ; Žerovnik J. ; Borštnar Kljajić M ; Drobne S. - Ljubljana : Slovensko društvo Informatika. Sekcija za operacijske raziskave, 2015, 29-34
ISBN
978-961-6165-45-7
Skup
International Symposium on Operations Research
Mjesto i datum
Bled, Slovenija, 23.09.2015. - 25.09.2015
Vrsta sudjelovanja
Predavanje
Vrsta recenzije
Međunarodna recenzija
Ključne riječi
decision support ; multi-criteria decision making ; utility function ; DEX ; UTADIS ; conjoint analysis ; direct marginals method
Sažetak
DEX is a qualitative multi-criteria decision analysis method. It provides support to decision makers in evaluating and choosing decision alternatives, using discrete attributes and rule- based utility functions. This work builds upon our previous attempt of approximating DEX utility functions with methods UTA and ACUTA, aimed at improving the sensitivity of qualitative models and providing an interpretation of DEX utility functions. In this work we empirically compare three methods for approximating qualitative DEX utility functions with piecewise-linear marginal utility functions: Direct marginals, UTADIS and Conjoint analysis. The results show that these methods can accurately approximate complete, monotone DEX utility functions.
Izvorni jezik
Engleski
Znanstvena područja
Računarstvo
