Expected value of distribution information for the newsvendor problem

This paper extends previous work on the distribution-free newsvendor problem, where only partial information about the demand distribution is available. More specifically, the analysis assumes that the demand distribution f belongs to a class of probability distribution functions (pdf) ℱ with mean μ and standard deviation σ. While previous work has examined the expected value of distribution information (EVDI) for a particular order quantity and a particular pdf f, this paper aims at computing the maximum EVDI over all f∈ℱ for any order quantity. In addition, an optimization procedure is provided to calculate the order quantity that minimizes the maximum EVDI.