Description:
The P-model objective of chance-constrained programming reflects the desire of management to maximize the probability of achieving or exceeding a given level of performance. This paper explores the implications of being able to make a sequence of decisions with the P-model objective function, and introduces some new possibilities for the P-model objective function that arise from the sequential nature of the problem. However, it is shown that, except for a special form of the objective function, the maximization in the sequential problem is not of a quasiconcave function, so that local optimality conditions are not sufficient to guarantee that a proposed solution is a global optimum. An example is worked out in detail to illustrate the computations involved for the objectives considered.