Differentiated service inventory optimization using nested partitions and MOCBA

In this paper, we consider a differentiated service inventory problem with multiple demand classes. Given that the demand from each class is stochastic, we apply a continuous review policy with dynamic threshold curves to provide differentiated services to the demand classes in order to optimize both the cost and the service level. The difficult features associated with the problem are the huge search space, the multi-objective problem which requires finding a non-dominated set of solutions and the accuracy in estimating the parameters. To address the above issues, we propose an approach that uses simulation to estimate the performance, nested partitions (NP) method to search for promising solutions, and multi-objective optimal computing budget allocation (MOCBA) algorithm to identify the non-dominated solutions and to efficiently allocate the simulation budget. Some computational experiments are carried out to test the effectiveness and performance of the proposed solution framework.