Bellingham Symposium on Modeling and Data Analytics 2021
Sponsored by the Center of Operations Research and Management Sciences at Western Washington University
The first Bellingham Symposium on Modeling and Data Analytics (BSOMADA) sponsored by Center for Operations Research and Management Science (CORMS) at WWU will be held from August 30 to August 31. The symposium will cover recent developments in quantitative modeling and data analytics. Due to the COVID-19 pandemic, this year's symposium is virtual (via ZOOM). The symposium will feature six invited talks given by six excellent researchers in the areas of quantitative modeling and data analytics.
If you are interested in attending this event, please contact Dr. George Zhang at george.zhang@wwu.edu or Dr. Xiaofeng Chen at chenx@wwu.edu
Data-Driven Platelet Inventory Management
Dr. Douglas Down (McMaster University, Canada)
Abstract
Platelet products are both expensive and have very short shelf lives. In addition, usage rates are highly variable, complicating inventory management for hospitals. Using a large clinical dataset (containing daily platelet transfusions and additional features including laboratory test data) from four hospitals in Hamilton, Ontario, Canada, we discuss two related problems. The first is developing an efficient model for forecasting demand. We discuss the effectiveness of various methods, including time series and machine learning approaches and provide several recommendations. The resulting forecast models still exhibit errors, so the challenge is how to effectively incorporate the forecasts for inventory management. We present two data-driven ordering schemes that, in simulations, outperform current ordering practice.
Bio
Douglas Down is with the Department of Computing and Software at McMaster University, where he also serves as the Academic Director of the Computing Infrastructure Research Centre. He obtained his PhD from the University of Illinois at Urbana-Champaign and was previously a faculty member at the Georgia Institute of Technology. His research interests include queuing theory, data center operations, and scheduling/inventory management with predictions.
A Multi-period Newsvendor Model under the Distortion Risk Measure
Dr. Hui Shao (Zhejiang University, China)
Abstract
The classic risk-neutral problem is to find the order quantity maximizing the one-period expected profit. This paper tries to extend the model in the following aspects: 1) We consider a risk-averse newsvendor under the distortion risk measure, and seek optimal solutions for a trade-off between return and risk. 2) In addition to selling newspapers, we assume that the newsvendor also invests in risky assets (e.g., stocks). To this end, a multi-period optimization model is proposed for the newsvendor to decide the optimal strategy. This paper is at a very early stage, and comments are greatly appreciated.
Bio
Dr. Hui Shao joined the International Business School at the Zhejiang University as an Assistant Professor in November 2020. He received Ph.D. in Quantitative Finance from Peking University (2017). From 2015 to 2017, he was a research assistant at Center of Quantitative Finance of NUS, and from 2017 to 2020, he was a research fellow at Risk Management Institute of National University of Singapore. Dr. Shao's research interests include derivative pricing, portfolio selection, and FinTech, using mathematical tools such as stochastic control, convex optimization, and numerical solution of partial differential equations.
Delay Information on A Ridesharing Platform: Large-Scale Field Experiments
Dr. Yong-Pin Zhou (UW Seattle)
Abstract
We study how to use delay information to better match the supply with demand on ride-sharing platforms through large-scale field experiments. We design randomized field experiments to examine how consumers’ abandonment behavior responds to delay information displayed on their app, which includes the initial magnitude of the delay estimate, how it is updated over time, and its granularity. To the best of our knowledge, we are the first to empirically study the impact of delay information on rides-sharing platforms. Our empirical results provide insights on how the platform can use the delay information to improve customer experience and better match supply and demand. Our insights have been validated by the ride-sharing platform we partnered with, and our recommendations have been implemented on the ride-sharing platform nationally.
Bio
Yong-Pin Zhou received his M.S. and Ph.D. in Managerial Science and Applied Economics from the Wharton School at University of Pennsylvania. He is a Professor of Operations Management and Ever McCabe Fellow at the Foster School of Business of the University of Washington, Seattle since 2007.Yong-Pin Zhou’s main research interests are in the areas of service operations management, supply chain management, and data-driven operations. Furthermore, Yong-Pin Zhou has done research in the interface areas between operations management and marketing, and between operations management and information systems. Yong-Pin Zhou’s research has been published in the leading research journals in Operations Management.
Yong-Pin Zhou is a recipient of the NSF Faculty Early Career Development (CAREER) Award, and has been selected as Welliver Faculty Fellow by Boeing in Summer 2009. Yong-Pin Zhou is actively involved in the wider academic community and served as an Associate Editor for Management Science, Manufacturing & Service Operations Management, and Operations Research.
Effects of Waiting Time Information on the Performance of a Two-Tier Service System
Dr. Ilhyung Kim/Dr. Mark Springer (Western Washington University)
Abstract
Despite free health care services in Canada, many Canadians choose to cross the U.S. border and pay for operations at U.S. hospitals. Shorter and more predictable waiting times are an important factor driving these decisions. Most Canadian patients, especially when their conditions are not life threatening, will be placed on a waiting list; this may require them to wait for months or even years. In addition, while Canadian patients only receive estimates of the average waiting time for their Canadian healthcare service, U.S. services generally provide a specific service date based on real-time information. This situation motivated us to consider a two-tier service system where one provider charges a premium or “toll” relative to the other service and where the two service providers may provide different levels of information to their customers. We examine how such differential information sharing affects the performance of the systems. We find that “more isn’t always better”; while the sharing of real-time information by the toll service provider is beneficial to both service providers, the sharing by the non-toll or “free” service provider is detrimental not only to the toll service provider but may also, under certain circumstances, harm the free service provider as well.
Bios
Ilhyung Kim is a Professor in the Department of Decision Sciences at Western Washington University. He holds a PhD in Operations and Technology Management from the Anderson Graduate School of Management at the University of California, Los Angeles (UCLA). His current research interests include supply chain coordination, manufacturing and service systems design, and statistical learning. He teaches business statistics and predictive analytics.
Professor Springer teaches courses on management science and quality management. His current research interests include two-tiered service systems, learning theory, and e-service quality. He research has appeared in the Journal of Operations Management, the European Journal of Operational Research, the International Journal of Production Research, and other outlets
Risk-Sensitive Markov Decision Processes and Reinforcement Learning
Dr. Li Xia (Sun Yat-Sen University,, China)
Abstract
Reinforcement learning (RL) has been receiving intensive research attention since the significant success of AlphaGo. Markov decision processes (MDP) are used as the mathematical model of RL. However, most of the current RL and MDP studies focus on the optimization objective of cumulative discounted rewards. Risk-related objectives are also important for many practical systems, such as the risk management in finance. In this talk, we will introduce some theoretical results on the long-run variance optimization problems in the framework of MDPs, where variance is a widely used metric for measuring risk. Because of the quadratic form of variance function, the long-run variance cost function depends on the whole policy and is not Markovian. The long-run variance optimization problem is not a standard MDP model and the classical Bellman optimality equation does not hold. We study this problem from a new perspective called the sensitivity-based optimization, which help us derive some new advances: by defining a pseudo-variance quantity, we derive a variance difference formula which has an elegant form to quantify the difference of long-run variances under any two policies. Based on the variance difference formula, we obtain a necessary and sufficient condition for the local optimum in the mixed policy space, which is only a necessary condition for the global optimum. We further derive the so-called Bellman local optimality equation and the optimality of deterministic policies is also proved. We further develop a policy iteration type algorithm to minimize the long-run variance and its local convergence is also proved. Finally, we extend the theoretical results to RL algorithmic studies and apply them to several practical problems, including the power fluctuation reduction of wind and battery storage system and the portfolio management problem in financial engineering. The latest research discovers that our work is able to extend to MDPs with CVaR (Conditional Value-at-Risk) objectives.
Bio
Dr. Li Xia is a professor with the Business School, Sun Yat-Sen University, Guangzhou, China. He received the Bachelor and the Ph.D. degree in control theory both from Tsinghua University, Beijing, China, in 2002 and 2007, respectively. After PhD graduation, he worked at IBM Research China and the King Abdullah University of Science and Technology (KAUST) Saudi Arabia. Then he returned to Tsinghua University as a faculty since 2011. In 2019, he joined Sun Yat-Sen University as a full professor. He was a visiting scholar at Stanford University, the Hong Kong University of Science and Technology, etc. He serves as an associate editor of IEEE Transactions on Automation Science and Engineering, Discrete Event Dynamic Systems, etc. His research interests include the methodology research in Markov decision processes, reinforcement learning, queuing theory, and the application research in energy systems, financial technology, etc.
Recommendations for Designing and Publishing Mixed-Methods Research
Dr. Christopher B. Califf (Western Washington University)
Abstract
This talk provides an overview of mixed-methods research, and reviews several important guidelines that authors should address when designing and publishing mixed-methods research. I use my own published mixed-methods study, “The Bright and Dark Sides of Technostress: A Mixed-Methods Study Involving Healthcare IT,” as an example of how to instantiate the recommended guidelines.
Bio
Christopher B. Califf (califfc@wwu.edu) is an associate professor of Information Systems in the Department of Decision Sciences at Western Washington University. Chris’ research focuses on a variety of topics that cover the interplay between technology and people including technology-induced stress (technostress), healthcare IT, mixed-methods and qualitative research methodologies, trust in technology, and cloud computing
His research has been published in outlets such as MIS Quarterly, Journal of Information Technology, MIS Quarterly Executive, AIS Transactions on Human-Computer Interaction, and LSE Business Review, among others. Chris serves as an associate editor for several IS conference proceedings and was recently a co-editor for a special issue of Health Policy & Technology. Chris also reviews for many journals including MIS Quarterly and Information Systems Research.