All attendees receive free access to the INFORMS 2020 TutORials in Operations Research online content concurrently with the meeting. Registrants of the 2020 INFORMS Annual Meeting have online access to the 2020 chapters, written by select presenters, beginning on November 8, 2020. Access this content using the link provided to all attendees by email or, if you are a 2020 member, simply login to INFORMS PubsOnLine.
The TutORials in Operations Research series is published annually by INFORMS as an introduction to emerging and classical subfields of operations research and management science. These chapters are designed to be accessible for all constituents of the INFORMS community, including current students, practitioners, faculty, and researchers. The publication allows readers to keep pace with new developments in the field and serves as augmenting material for a selection of the tutorial presentations offered at the INFORMS Annual Meetings.
A Framework for the Management of Innovation
Innovation is critical in all types of organizations. Yet, it remains one of the most challenging efforts to perform and manage. This presentation provides a conceptual framework that breaks the complexity of innovation efforts down into three key phases, namely ideation, selection and execution. We then review the extant literature that has emerged on the management of innovation and we identify that there is no single standard recipe that enables effective innovation management.
Authors: Stelios Kavadias; Jeremy Hutchison-Krupat, University of Cambridge
AI and Algorithmic Bias: Source, Detection, Mitigation and Implications
Artificial intelligence (AI) and machine learning (ML) algorithms are widely used throughout our economy in making decisions that have far-reaching impacts on employment, education, access to credit, and other areas. Initially considered neutral and fair, ML algorithms have recently been found increasingly biased, creating and perpetuating structural inequalities in society. With the rising concerns about algorithmic bias, a growing body of literature attempts to understand and resolve the issue of algorithmic bias. In this tutorial, we discuss five important aspects of algorithmic bias. We start with its definition and the notions of fairness policy makers, practitioners, and academic researchers have used and proposed. Next, we note the challenges in identifying and detecting algorithmic bias given the observed decision outcome, and we describe methods for bias detection. We then explain the potential sources of algorithmic bias and review several bias-correction methods. Finally, we discuss how agents’ strategic behavior may lead to biased societal outcomes, even when the algorithm itself is unbiased. We conclude by discussing open questions and future research directions.
Authors: Yan Huang; Param Vir Singh, Runshan Fu, Carnegie Mellon University
Analytical OppORtunities in Tax Administration
The mission of tax administration includes providing service to assist taxpayers with their tax-related filing, reporting, and payment obligations; enforcing tax laws; and adjudicating any tax issues. Taxpayers have an expectation that these activities are administered in a fair and impartial manner with effective stewardship of tax dollars. To deliver timely and accurate service, multiple channels are needed to meet a diverse set of service preferences, including phone, in-person, internet, and written correspondence. Optimizing service delivery objectives across channels involves balancing the demand for service against internal capacity constraints, and through external partnerships with tax preparers, financial institutions, volunteer groups, and other stakeholders. Enforcing tax laws, whether civil or criminal, requires identifying compliance risk in areas that include ID theft, refund fraud, multi-party tax schemes, off-shore transactions, and money laundering. It requires optimizing workload decisions across business processes, often interconnected, in examination, collection, and criminal investigation domains, to achieve the best set of outcomes for both IRS and taxpayers. To address these challenges, modern tax administration has an opportunity to leverage state-of-the-art analytics in areas such as machine learning, deep learning, natural language processing, graph mining, and simulation to improve decision-making and meet mission objectives. This tutorial will highlight how the Internal Revenue Service is using these capabilities to transform operations and create a more efficient and cost-effective tax administration.
Author: Ron Hodge, Jeff Butler, Michael Dunn, IRS
Diversity, Equity and Inclusion in the Decision Sciences: A Research Agenda for Scholarship, Practice and Service
Diversity, equity and inclusion (DEI) refer to strategies and processes that enable organizations to become more reflective of and responsive to identities, values and experiences of different stakeholder groups. Organizations committed to DEI may better achieve their own missions and improve the well-being of their stakeholder groups and of society at large. In this tutorial, I will explore the presence of DEI in the professional practice of OR/analytics, examine the role of DEI in OR/analytics scholarship, and present an agenda for the increased application of DEI principles in the OR/analytics profession. Given increased attention to social and racial justice, this work will demonstrate how OR/analytics can use DEI principles to enable our society become more equitable, more just, and more welcoming of people from diverse backgrounds, identities and communities.
Author: Michael Johnson, University of Massachusetts Boston
Machine Learning in Health Care: Fairness, Issues, and Challenges
This tutorial discusses considerations relevant to operations researchers undertaking machine learning projects in the healthcare domain. We introduce readers to the unique considerations of healthcare data, from data cleaning and preparation to feature generation and dimension reduction. We then focus on the modeling techniques that tend to perform well in the healthcare context, and we highlight common stumbling blocks. We close with a discussion of fairness and transparency in healthcare modeling. This tutorial assumes that readers are familiar with basic machine learning techniques and terminologies but do not have a background in the health care field.
Authors: Margrét Bjarnadóttir, University of Maryland, David Anderson (Villanova University)
Information Design in Operations
Consider a set of agents (receivers) whose payoﬀs depend on an underlying state of the world as well as each other’s actions. Suppose that a designer (sender) commits to a signaling mechanism which reveals payoﬀ-relevant signals to agents when the state is realized. The availability of such signals inﬂuences the agents’ actions, and by choosing the signaling mechanism appropriately the designer can induce a desired outcome. Information design studies signaling mechanisms that maximize the payoﬀ of the designer. In this paper, we ﬁrst present the classical information design framework and discuss diﬀerent approaches for characterizing the optimal information structures. We then discuss various applications in the recent operations literature. The applications include signaling (i) content/product quality in networked systems, (ii) product availability in revenue management settings, and (iii) seller quality in twosided markets. Finally, we present recent work that discusses the design of optimal information structures when some of the key assumptions in the classical information design problems (which may not hold in operational settings of interest) are relaxed.
Author: Ozan Candogan, University of Chicago
Military Operations Research: A Career’s Worth of Opportunities in the United States Army
Military Operations Research offers a breadth of opportunities for the study and practice of operations research addressing both military and security issues. I have served for over twenty years as an Army Operations Research and Systems Analyst and will share some thoughts on the field and its opportunities. I will share some recent research on military operations research to explore the breadth of opportunities to apply operations research to military problems. I provide three examples of work that I have done that showcase the increasing complexity level of problems within military operations research that correspond with the career development of an analyst.
Author: Andrew Hall, West Point
On the Use of Operations Research and Management in Public Education Systems
The tutorial explores questions related to the use of OR/MS to improve public education systems. Drawing on papers from the literature and experiences of researchers and practitioners, we answer these questions and highlight emerging research areas. We trace the history of this research with events that changed public educations.
Authors: Karen Smilowitz, Northwestern University; Samantha Keppler, University of Michigan
Clearing Financial Networks: Impacts on Equilibrium Asset Prices and Seniority of Claims
Clearinghouses have been mandated as de facto central nodes in financial networks. Other services aim at solving various network problems such as reducing cycles. All these services face complex challenges, ranging from the direct effect on the network topology to the indirect effects on asset prices and on the end users. In this tutorial, we offer an introduction to the quantitative modeling of clearing systems. We analyze joint equilibria for the network payments and the asset prices. We raise two main points that have received less attention in the literature. The first point is that an equilibrium price for the asset prices may not be achieved in a general payment network. The second point is that portfolio compression and clearinghouses modify the seniority structure in the network, and end users that are not part of multilateral clearing arrangements become junior.
Authors: Andreea Minca, Cornell University; Hamed Amini, Georgia State University
Stochastic Market Microstructure Models of Limit Order Books
Many financial markets are operated as electronic limit order books (LOB). Over short time scales, seconds to minutes, LOBs can be best understood and modeled as stochastic dynamical systems, and, specifically, ones that exhibit interesting and relevant queueing phenomena. I will offer a brief overview of algorithmic trading in a limit order book, and highlight how queueing phenomena play an important role in trade execution, and as a consequence in market behavior.
Authors: Costis Maglaras, Columbia University; Rama Cont, University of Oxford
In this tutorial, we describe predictive and prescriptive analytical methods that assist primary enterprises in the wine supply chain in their decision-making processes. The tutorial begins with predictive models that estimate the true value of wine futures prices. These estimation models are essential to the financial exchange called the London International Vintners Exchange, Liv-ex, where wine futures contracts are traded. Coined as “realistic prices” by Liv-ex, these predictive models assist buyers in their purchasing decisions as they can determine whether a futures contract is underpriced or overpriced. The tutorial then develops risk mitigation models to assist winemakers so that they battle with uncertainty in weather conditions and tasting expert reviews. These prescriptive models rely on predictive analytics which help determine consumers’ utilities from buying the wine in advance, or later, or not purchasing at all. Prescriptive models such as a multinomial logit model focus on determining how much of the wine should be sold in advance in order to reduce the risk exposure and maximize the expected profits of the winemaker. On the buyer side, the tutorial introduces stochastic portfolio optimization models for wine distributors and importers in their decisions regarding how to allocate their limited budgets between wine futures contracts and bottled wine. These prescriptive models are, once again, built on predictive analytics that estimate the evolution of futures and bottle prices over time under fluctuating market and weather conditions.
Author: Burak Kazaz, Syracuse University