Optimization model and solution method for dynamically correlated two-product newsvendor problems based on Copula

Abstract

<p style='text-indent:20px;'>In this paper, a two-product newsvendor problem is taken into consideration, where the demands of products are correlated random variables and the buyer is risk-averse. Some important qualitative properties of the constructed model are analyzed, particularly the gradient information of the model is obtained and incorporated into solution method of the model. Based on the theory of Copulas, an efficient algorithm, called the feasible-direction based BFGS algorithm, is developed for solution of the constrained optimization model. Case study shows the efficiency of model and algorithm, and numerical results demonstrate that compared with the situation of independent demands, the total order quantity reduces against the high dynamic interrelation coefficient of two demands with the same degree of risk-aversion, and the optimal order quantity decreases as the degree of risk-aversion becomes greater.