Premium
Fuzzy versus stochastic approaches to multicriteria linear programming under uncertainty
Author(s) -
Slowinski R.,
Teghem J.
Publication year - 1988
Publication title -
naval research logistics (nrl)
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.665
H-Index - 68
eISSN - 1520-6750
pISSN - 0894-069X
DOI - 10.1002/1520-6750(198812)35:6<673::aid-nav3220350612>3.0.co;2-l
Subject(s) - linear programming , stochastic programming , fuzzy logic , mathematical optimization , computer science , mathematics , mathematical economics , econometrics , artificial intelligence
Recently, both authors independently proposed two different approaches to multicriteria linear programming under uncertainty with a view of an application to some long term planning problems. Slowinski [11] has developed a method called FLIP ( Fuzzy LI near P rogramming) based on the application of fuzzy numbers for modeling imprecise data. On the other hand, Teghem et al. [17] have proposed the method STRANGE ( STRA tegy for N uclear G eneration of E lectricity), a stochastic approach to the same problem. Both methods are interactive and at each step present to the decision maker (DM) a large representation of efficient solutions. The aim of this study is to compare FLIP and STRANGE. A didactic example is first defined and resolved by both methods. Next, every stage of both procedures is analyzed and compared on the basis of this example; taking into account imprecise data, formulation of deterministic multicriteria problems associated with the original problem, getting the first compromise solution, the role of the DM in the interactive decision‐making steps, etc. For each of these stages, possible limitations, advantages, and inconveniences of both methods are emphasized. General conclusions following from this comparison are finally drawn.