An Entropy Based Methodology for Valuation of Demand Uncertainty Reduction

Abstract

We propose a distribution-free entropy-based methodology to calculate the expected value of an uncertainty reduction effort and present our results within the context of reducing demand uncertainty. In contrast to existing techniques, the methodology requires neither sampled observations of demand nor a priori assumptions regarding the underlying demand distribution. Rather, leveraging the maximum entropy principle to assign a probability density over all possible demand distributions enables modeling of both one's present state of uncertainty and one's potential future states of uncertainty. We demonstrate that this probability assignment is intuitively satisfying, theoretically justified, and done in a manner that is completely consistent with a decision maker's current information (or lack thereof). Theoretical and numerical results for valuing uncertainty reductions without knowing an underlying demand distribution are explored and contribute to the existing distribution free literature. We leverage these results to answer an often overlooked question in demand management: "Is there value in further reducing my demand uncertainty or do I act on my currently available information?''