Monday, August 28

Monday, August 28 – Supply Chain Management

Workshop Chair: Anton Kleywegt

Time: 9:00 – 10:25
Speaker: Anton Kleywegt
Title: Revenue Management
Abstract: This tutorial gives an overview of revenue management models and methods. Some of the classical models and methods will be briefly discussed. Then we will cover some of the more modern models and methods, such as assortment optimization and pricing with various discrete choice models.

Time: 10:35 – 12:00
Speaker: He Wang
Title: Online Learning and Bandit Problems in Supply Chain and Revenue Management

Time: 12:00 – 13:30 – Lunch

Time: 13:30 – 14:00
Speaker: Satya Malladi
Title: Dynamic Mobile Capacity Logistics and Inventory Control
Abstract: We investigate the problem of dynamically planning inventory and capacity logistics for a mobile modular production system. Faced with state and action space explosion, we identify upper and lower bounds on the optimal cost function and propose rollout based heuristics that overcome the computational hurdle and determine near-optimal decisions. Through a systematic computational study, we quantify the value addition of mobile modular production systems over traditional fixed production systems.

Time: 14:00 – 15:00
Speaker: Roberto Pinto
Title: Rationing problems in the supply chain
Abstract: In many business scenarios where companies face uncertain demand and building extra capacity does not represent a viable option, the available production capacity or amount of stock might not be sufficient to fulfill every possible demand realization. Therefore, it is necessary to devise proper rules to distribute the availability and maximize the expected results. In this talk, the inventory rationing problem in a vertically integrated distribution system is discussed, providing solution procedures under different objectives.

Time: 15:15 – 16:00
Speaker: Hongzhang Shao
Title: Demand Estimation with Unobserved Stockouts

Time: 16:00 – 16:30
Speaker: Fabien Caspani and Jye-Chyi Lu
Title: Dynamic Category Planning With Forecast Updates and Space Constraint – Applications Include Pop-Up Store Assortment Decision
Abstract:In a pop-up store items sharing similar characteristics are grouped into categories. Retailers procurement decision for a given category is made once during the planning horizon for maximizing the expected profit from sales. The sequence of procurement decisions and space allocations for each category are not known in advance, and are dynamically made to benefit from forecast updates over the course of the planning horizon. The number of possible sequences exponentially increases with the number of categories. This study shares some similarities with the multiitem newsvendor problem where a capacity constraint is imposed (e.g., Turken et al., 2012), but extends their static decision process to a dynamic situation with updates of advance information about category demands. The study also extends separate optimizations about sequencing and allocation decisions (e.g., Wang et al, 2012; Papier, 2016) to a dynamic joint decision process. We propose a tractable solution method that first determines the sequence of procurement decisions made in the subsequent periods by using an approximation of the allocation problem. The optimal allocation is then found by using this sequence. Strength and weakness of a few developed algorithms are compared with a considerable number of simulation experiments designed for mimicking various real-life market conditions. Limited analytical studies provide partial-ordering of procurements for solving sequence decision problems.

Time: 16:30 – 17:00
Speaker: Yi Zhang