where here α is a pre-specified positive number?
ρ(û;α) = max min r(d,u) d∈D u∈U(α,û) But most of all, Info-Gap’s thesis that satisficing has an advantage over optimizing is, mathematically speaking, pointless, indeed counter-productive. For, it can be easily shown that any satisficing problem of the form r(d,u) ≥ r*, ∀u∈U(α,) can be re-formulated as an (equivalent) optimization problem such that any optimal solution to the latter is a feasible solution to the former and vice versa. So, the real issue here is not whether we should optimize or satisfice. Rather the real issue is the following. Given a state-of-affairs that we formulate as a decision-making problem, we consider the tasks, aims, and goals that the problem presents us with so that the question we ask ourselves is: “what should be optimized and what should be satisfied?” See my discussion on this topic. Remark: Info-Gap’s rhetoric on “satisficing” is astounding, especially when it is set off against its rhetoric on severe uncertainty. If you are unaware of what Info-Gap t
