Resource allocation under uncertainty: choice models and computational procedures

The deterministic and stochastic versions of the resource allocation problem have already been discussed in the literature. The goal of this contribution is to formulate optimization models applicable to the problem of resource allocation under uncertainty, which signifies that profits resulting from the assignment of a quantity of resource to a given activity, are defined as random variables with unknown distribution. The author presents four models depending on the attitude of the decision-maker towards states of nature that may occur, and refers to the rules formulated by Wald, Hurwicz, Bayes and Savage to this end. Possible computational procedures, allowing finding the optimal solution for each case, are also analyzed. Apart from the dynamic programming, two simplified methods used for the deterministic version of resource allocation can also be applied when decisions are made under uncertainty. However, these two methods require that the problem fulfil additional assumptions, which are partially different from those formulated for the deterministic approach.