Decision making problems in a general environment
1988; Elsevier BV; Volume: 25; Issue: 2 Linguagem: Inglês
10.1016/0165-0114(88)90182-0
ISSN1872-6801
AutoresManuel Jorge Bolaños, M. Teresa Lamata, Serafı́n Moral,
Tópico(s)Fuzzy Systems and Optimization
ResumoA general model for decision problems is presented, based on the consideration of an evidence given by a basic probability assignment, which is the information about states of nature. Decision rules are defined in two steps by means of a composition of functions. Firstly, the lower and upper expected values provide limits to the expected utility value for each decision. The function of the second step reflects the decision maker's subjectivity and lets us model it between these two boundaries by means of a convex combination. The relationships between evidences and fuzzy measures are shown in Section 2 and some relevants properties of lower and upper expected values are proved in Section 3. In Section 4, a definition of inclusion relation on the set of evidence is used to study existing relations among results which can be obtained from different basic probability assignments, and, finally, we show the way in which our model includes the most classical criteria, such as min-max or mathematical expectation criteria in uncertainty or risk environments.
Referência(s)