Poster Session and Coffee
Tuesday, December 4, 2018 - 2:30pm - 4:00pm
- Manufacturers' Assortment Planning and Pricing in A Competitive Two-Tier Supply Chain
Juan Xu (University of Illinois at Urbana-Champaign)
We consider assortment competition in a two-tier supply chain with multiple manufacturers and a single wholesaler. Each manufacturer chooses an assortment to supply to the wholesaler, and the latter sets market prices on these items and sells them to consumers. We consider both the case where the wholesale prices are fixed exogenously and the case where they are chosen by manufacturers to maximize their own profits. The division of manufacturers' market shares depends on the consumer choice formulated by the standard Multinomial Logit (MNL) model. By exploring the connection between manufacturers' profit functions with the classical assortment optimization problem with the MNL model, we are able to characterize the structure of a manufacturer's optimal response to other manufacturers' assortment and/or pricing decisions. Based on this characterization, we establish the existence of an equilibrium of the competition. Furthermore, we apply our results to a special case with two manufacturers, each of which has two products that differ by their margins and attraction factors to manufacturers' direct buyers. We show that in this competition, under a certain condition, each manufacturer is more likely to provide an assortment with a larger variety in a two-tier supply chain.
- Box Suite Recommendation
Stuart Rogers (University of Minnesota, Twin Cities)
E-commerce retailers such as Amazon, Walmart, and Target ship millions of orders to customers every day. Unfortunately, for some retailers, the boxes used to ship these orders are on average 50% or more empty! Choosing an optimal suite of boxes for shipping all these orders will minimize the volume of empty space shipped, thereby saving millions of dollars per year in shipping and material costs and reducing the carbon footprint of online retail. This poster presents an algorithm for recommending an optimal box suite based on a large set of historical customer orders.
- Robust Multi-product Newsvendor Model with Substitution under Cardinality-constrained Uncertainty Set
Jie Zhang (Virginia Polytechnic Institute and State University)
This paper studies Robust Multi-product Newsvendor Model with Substitution (R-MNMS), where the demand is stochastic and is subject to cardinality-constrained uncertainty set. The goal of this work is to determine the optimal order quantities of multiple products to maximize the worst-case total profit. To achieve this, we first show that for given order quantities, computing the worst-case total profit in general is NP-hard. Therefore, we derive the closed-form optimal solutions for the following three special cases: (1) if there are only two products, (2) if there is no substitution among different products, and (3) if the budget of uncertainty is equal to the number of products. For a general R-MNMS, we formulate it as a mixed integer linear program with an exponential number of constraints, and develop a branch and cut algorithm to solve it. For large-scale problem instances, we further propose a conservative approximation of R-MNMS and prove that under some certain conditions, this conservative approximation yields an exact optimal solution to R-MNMS. The numerical study demonstrates the effectiveness of the proposed approaches and the robustness of our model.