Maxmin expected utility with non-unique prior (Q1122462)

An axiomatic foundation of the maxmin expected utility decision rule is proposed here. The main difference among the model of \textit{F. J. Anscombe} and \textit{R. J. Aumann} [Ann. Math. Stat. 34, 199--205 (1963; Zbl 0114.07204)] and the present one lies in the phrasing of the independence axiom. Uncertainty aversion is additionally assumed. The considered model extends the classical expected utility but it cannot be represented in the non-additive expected utility theory. A concept of independence in the case of non-unique prior is proposed.