O.R. and M.S. Revisited in the Case of Uncertain and Subjective Data

Summary

Everybody who is confronted with modeling of micro or macro economical systems knows that classical deterministic or probabilistic data are more and more difficult to obtain. The environment changes too rapidly to obtain past data for the future, even the near future. In this case, expert advices, personal opinions, simulations with computers, various investigations and so on can be acceptable. But this kind of data are, in their nature, uncertain and subjective. Of course, objective data are ideal and when measurement is possible it should be done. But for lack of such ideal informations, other less reliable informations, must be used.

When uncertain and subjective data must be introduced in economic models, other approaches than deterministic models are convenient: intervals of confidence, fuzzy subsets, probabilistic subsets, expertons, etc. The concept of possibility will take the place of the usual and stronger concept of probability. This paper will show how to specify and use these novel concepts in O.R. and M.S. modeling.