Article
Engineering, Manufacturing
Jian Yang, Jim (Junmin) Shi
Summary: This article studies inventory control in the presence of lost sales and censored demand. The demand distribution is largely unknown in a long-run average framework. As long as the stationary inventory costs are strictly convex, with the second lost item costing more than the first one, the regret would be Omega(T-2/3). The article proposes a policy that orders high levels during designated learning periods and uses base-stock levels tailored to near-empirical distributions formed over the learning periods during the remaining doing periods. This policy can achieve a matching O(T-2/3) upper bound even for nonperishable items. Numerical experiments illustrate the competitiveness of this separate learning-doing policy.
PRODUCTION AND OPERATIONS MANAGEMENT
(2023)
Article
Mathematics, Applied
Khwanchai Kunwai, Fubao Xi, George Yin, Chao Zhu
Summary: Motivated by applications in natural resource management, risk management, and finance, this paper focuses on an ergodic two-sided singular control problem for a general one-dimensional diffusion process. The optimal reward value and control policy are derived using the vanishing discount method under mild conditions. Additionally, the Abelian and Cesaro limits are established, and a direct solution approach is provided.
APPLIED MATHEMATICS AND OPTIMIZATION
(2022)
Article
Management
Ruud H. Teunter, Stefan Kuipers
Summary: This study examines the optimal inventory control of two products with demand substitution. Using a simplified Economic Order Quantity model, the authors present new insights into the optimal ordering strategies for two substitute products and find that partial substitution is always achieved through one-way substitution rather than two-way substitution.
OMEGA-INTERNATIONAL JOURNAL OF MANAGEMENT SCIENCE
(2022)
Article
Automation & Control Systems
Jianguo Zhao, Chunyu Yang, Weinan Gao, Hamidreza Modares, Xinkai Chen, Wei Dai
Summary: This paper addresses the issue of linear quadratic tracking control (LQTC) with a discounted cost function for unknown systems. Existing design methods often require a small discount factor for closed-loop stability, but solving the discounted algebraic Riccati equation may lead to ill-conditioned numerical problems with a small discount factor. By using singular perturbation theory, the full-order discounted Riccati equation is decomposed into a reduced-order Riccati equation and a Sylvester equation, allowing for the design of feedback and feedforward control gains. The resulting controller is proven to be stabilizing and near-optimal in solving the original LQTC problem. In the framework of reinforcement learning, on-policy and off-policy two-phase learning algorithms are derived for designing a near-optimal tracking control policy without prior knowledge of the discount factor. Comparative simulation results are provided to demonstrate the advantages of the proposed approach.
Article
Engineering, Manufacturing
Boxiao Chen
Summary: In a shifting demand environment, detecting and learning demand distributions solely from historical sales data is necessary, with the need for active exploration in inventory space for a reasonable algorithm. Theoretical lower bound is provided, showing that nonparametric learning algorithm can achieve convergence rate matching the bound.
PRODUCTION AND OPERATIONS MANAGEMENT
(2021)
Article
Engineering, Manufacturing
Zhaolin Li, Samuel N. Kirshner
Summary: This study focuses on contract design in a salesforce principal-agent model, revealing the critical role of variance information and the necessity of using distribution-free contracts to achieve the optimal outcome. The index of dispersion determines whether the optimal contract is linear or quadratic.
PRODUCTION AND OPERATIONS MANAGEMENT
(2021)
Article
Management
Ming Hu, Yan Liu
Summary: This study examines the competition between two platforms (such as Uber and Lyft) in a two-sided market and investigates the impact of various precommitments on equilibrium outcomes. The results show that commission precommitment is less profitable than no precommitment when the competition intensities of both sides are close. However, capacity precommitment leads to the most profitable outcome. Flexible competition modes benefit from higher demand uncertainty.
M&SOM-MANUFACTURING & SERVICE OPERATIONS MANAGEMENT
(2023)
Article
Operations Research & Management Science
Francisco Salas-Molina, David Pla-Santamaria, Juan A. Rodriguez-Aguilar
Summary: Cash managers rely on biobjective optimization models to select the best policies based on their risk preferences. However, the cash management literature lacks an analytical derivation of the efficient frontier, which would help cash managers evaluate the implications of selecting policies and risk measures.
ANNALS OF OPERATIONS RESEARCH
(2023)
Article
Computer Science, Artificial Intelligence
Dongping Pu, Guanghui Yuan
Summary: This paper proposes a two-sided matching model that considers the demands of subjects and the intermediary, and solves the complex matching demands through generalized preference order determination methods and common expected preference order algorithm. It provides a stable matching optimization model that can effectively solve the practical matching problem.
EXPERT SYSTEMS WITH APPLICATIONS
(2023)
Article
Engineering, Electrical & Electronic
Zhibin Zhao, Shibin Wang, David Wong, Chuang Sun, Ruqiang Yan, Xuefeng Chen
Summary: This paper proposes a robust enhanced trend filtering algorithm called RobustETF, which can extract trends in the presence of various types of non-Gaussian noise or outliers. The algorithm utilizes an extended EM algorithm to solve the resulting non-convex optimization problem.
Article
Management
Dennis Prak, Ruud Teunter, Mohamed Zied Babaic, John E. Boyland, Aris Syntetos
Summary: Existing methods for estimating parameters in inventory control lack guidance, with traditional MM and ML estimators proving to be less robust. We propose a new MM alternative that outperforms standard methods in accuracy and performance.
OMEGA-INTERNATIONAL JOURNAL OF MANAGEMENT SCIENCE
(2021)
Article
Management
Mehdi Davoodi, Michael N. Katehakis, Jian Yang
Summary: This article investigates a new model for dynamic inventory control problems, allowing decisions to be made in situations where the demand distribution is unknown. A strategy to control the growth of regret is proposed with the aim of addressing the long-term regret growth resulting from learning demand and making ordering decisions.
OPERATIONS RESEARCH
(2022)
Article
Engineering, Manufacturing
Xiangyu Gao, Huanan Zhang
Summary: This paper investigates a periodic-review multiproduct inventory system in which customers' purchasing decisions are influenced by product availabilities. A UCB-based learning framework is proposed to address the learning problem by utilizing sales information based on two improvement ideas. Improved UCB algorithms are developed for two specific systems with tight worst-case convergence rates. Extensive numerical experiments demonstrate the efficiency of the improved UCB algorithms.
PRODUCTION AND OPERATIONS MANAGEMENT
(2022)
Article
Engineering, Industrial
Julia Miyaoka, Katy S. Azoury
Summary: This article discusses a production-inventory system that switches production rates based on demand. It introduces a two-critical number policy for controlling the system and derives the steady-state distribution of the inventory level using a level crossing approach. The total expected costs for two different scenarios are calculated, and simplified solutions are proposed.
INTERNATIONAL JOURNAL OF PRODUCTION RESEARCH
(2023)
Article
Management
Recep Yusuf Bekci, Mehmet Gumus, Sentao Miao
Summary: Motivated by collaboration with a major fast-fashion retailer in Europe, this study focuses on the one-warehouse multistore (OWMS) problem in a two-echelon inventory control system when the demand distribution is unknown. The goal is to minimize the total expected cost, considering various cost factors. The challenge lies in dealing with censored demand and generating unbiased demand estimation. To address this, a primal-dual algorithm is proposed, which continuously learns the demand and makes inventory control decisions. The approach shows promising theoretical and empirical performances.
OPERATIONS RESEARCH
(2023)