OR/MS Tomorrow
Share:
Student Writing Competition !!!
Fall/Winter 2019 Issue: Operations Research and Game Theory
The online student membership magazine for INFORMS. The bi-annual publication provides a look at operations research and management science from the perspective of young people in those fields. Edited by a team of students and junior faculty, the magazine is written for students and aims to introduce topics relevant to them, highlight their accomplishments, and promote awareness of current events and issues in OR and MS.
A Brief Overview of Game Theory, OR, and Their Roles in Better Decision-Making
Operations Research (OR) is the scientific study of the management of operations and processes for the purpose of better problem-solving and decision-making (Horner, 2015). Using the tools of mathematics, statistics and computer science, OR researchers and practitioners are concerned with how managerial decisions that control the operations of the system of interest should be made and implemented to improve the targeted outcome(s) (Tanenbaum & Eilon, 2018). Some of the basic concepts of OR are as follows:
Bi-level programming with applications in engineering and economics
Have you ever played Texas hold’em? If you have, you probably realize that the dealer has an advantage: she gets to act last. This allows the dealer to see what everyone else at the table is betting and then decide on what she wants to do. Having seen how everyone else has acted, the dealer can make a more informed decision. So, based on the rules of the game, or the sequence in which information unfolds, a player can gain an advantage over other participants. In fact, it is the information structure of the game (in this case, Texas hold’em) that provides an advantage for the dealer. And this is precisely why the role of the dealer rotates from one player to another, one hand at a time, so that no one player remains in a permanently advantageous position.
Game Theory and Reinforcement Learning
Reinforcement Learning (RL) and Game Theory are two streams of mathematics with significant applications in solving real-life problems. Despite different origins, these methods share common traits in how the problems are defined in the game environment; i.e. states, agents and strategies (or policies). Reinforcement learning, a field of machine learning, is a ‘trial and error’ algorithm. Based on its observations, in RL an agent acts on an unknown environment to maximize the reward.
Game Theory and Machine Learning
With the onset of the New Year, it makes sense to review what has been accomplished in the past year; here, I briefly review three papers on machine learning and game theory. Many machine learning methodologies have been explored by either casting them into a game-theoretic framework or by using game theory on top of the existing machine learning framework. Popular examples include Generative Adversarial Nets (GANs) (Goodfellow et al., 2014) which correspond to a minimax two-player game between the generator and discriminator networks, hard margin support vector machine which can be modeled as a two-player zero-sum game (Aiolli et al., 2008), linear regression being modeled as a non-cooperative game (Ioannidis and Loiseau, 2013), and Adaboost (Freund and Schapire, 1997) which uses game-theory for online learning.
Application of Game Theory in Distributed Systems: An Interview with Dr. Daniel Grosu
Dr. Daniel Grosu is an Associate Professor of Computer Science and Director of the Parallel and Distributed Computing Lab at Wayne State University. Dr. Grosu and his team have published dozens of articles in top-tier peer-reviewed conferences and journals in the area of resource provisioning and pricing mechanisms in cloud, mobile edge, and vehicular edge computing systems. A main research focus of Dr. Grosu's team is the application of game theory in distributed systems, which is the subject of this interview.
Fair Division Problems: How to Cut Your Cake and Eat It Too
If there is a commodity that needs to be divided within a group of people, often times, this seemingly simple problem isn’t an easy task. Each person may have a different preference on different parts of the commodity, one person may not know another person’s preferences, the commodity may not even be divisible, and so on. This is a classical problem in game theory, aptly named fair division, which deals with keeping every person “happy” with their slices. From splitting rent between roommates when the rooms are of different sizes, to redrawing congressional districts for the nation’s top legislative body, fair division problems are ubiquitous.
Saving Costs While Saving Lives
“A new army is marching into the war against rising health care costs: engineer-mathematicians. These individuals occupy a field called operations research” wrote journalist Joel Shurkin (2013). With national discussions on healthcare topics such as the Affordable Care Act (ACA), opioid crisis, and drug pricing and shortages, it is clear that high costs and inefficient management of health care systems in the United States continue to be major concerns for government officials, policy makers and the general public.
OR/MS Tomorrow Spotlights
The chapter holds informative academic events, workshops, and social activities, which create opportunities for professional and personal development.
INFORMS Student Chapters: University of Pittsburgh, Auburn University, University of South Florida, University of Massachusetts Amherst and Mississippi State University
If your student chapter has not used the INFORMS Speakers Program before, this article is for you! Read on and find out about the fantastic benefits this program can provide for your chapter. If your chapter has used this program before, let this article serve as an inspiration to include a speaker in your upcoming events.