University of Wisconsin Madison College of Engineering

Decision Science and Operations Research


You are here:
  1. Industrial and Systems Engineering, ISyE > 
  2. Research > 
  3. Decision Science and Operations Research
Assistant professor Jim Luedke (left) and students

Assistant Professor James Luedtke (left) and students. 


The research and teaching program in decision science/operations research aims to improve the quality of decisions about managing scarce and valuable resources. Such resources include not only financial resources but also issues related to the quality of human life, medical treatment, the environment and many other important issues.


Faculty, DS/OR specialization

Jeffrey T. Linderoth — Convenor


Core faculty


Oguzhan Alagoz
Vicki Bier
Ananth Krishnamurthy
James Luedtke
Laura McLay
Leyuan Shi


Affiliate faculty


Michael Ferris
Christos Maravelias
Stephen Wright
Zhiguang (Peter) Qian


Emeriti faculty


Stephen Robinson
David Zimmerman



Goals of graduate study

Graduate study in decision science and operations research entails developing problem-solving skills that can be used to make and implement decisions as efficiently and effectively as possible. These problem-solving skills involve recognizing and identifying decision problems, as well as generating, evaluating, choosing and implementing solutions to these problems.



Students are encouraged to develop expertise in an applied area as well as to understand the theories and tools. Interests of the faculty in decision science/operations research include:


  • Manufacturing process optimization
  • Supply chain design and optimization
  • Medical decision making
  • Optimization of health-care systems
  • Performance analysis of computer systems
  • Security and critical-infrastructure protection
  • Modeling and simulation in defense analysis


Program requirements

Advising information [PDF]



Employment prospects

Decision science and operations research techniques are frequently used in jobs in areas such as:


  • Consulting companies
  • Software companies
  • In-house decision science and operations research groups of major corporations (e.g., airlines, computer companies, telecommunication firms)
  • The public sector (e.g., policy analysis, military operations and logistics)
  • Health care (e.g., patient-flow analysis, continuity of care across providers)
  • Industrial-research laboratories
  • National laboratories


In addition, PhD graduates in decision science/operations research also go on to faculty positions in industrial engineering departments, business schools, and schools of public policy. You can find more information on careers in decision science and operations research at:



Comments from recent graduates

“The Decision Science/Operations Research group provides a strong theoretical foundation, thus enabling students to apply OR principles to a variety of practical situations. ... The faculty and department staff members are always willing to help interested students.” — Product-launch analyst at a manufacturer of orthopedic implants


“If you thought courses were tough enough, wait till you figure out the impossibly long list of extra-curricular activities that you can plunge into! UW-Madison is a great place to build new friendships with peers as well as faculty.” — Business analyst at a credit-card company


MSIE, Decision Science/Operations Research specialization

Decision science/Operations Research brochure 2012 [PDF]


The MS program with specialization in decision science/operations research is designed to provide both balance and breadth in the student's understanding of decision science and operations research techniques and applications. To accomplish this, students must take at least two classes in optimization, at least two classes in stochastic processes, at least one class in simulation, and at least one class in the area of organizations, decisions, and implementation issues. The program is rounded out with electives, selected with the approval of the student's adviser. Flexibility is built into the program to accommodate a wide range of interests and applications.


PhD concentration areas

Doctoral students in decision science/operations research concentrate in one or more of three main areas:


Stochastic Processes and Simulation

Uncertainty pervades practical decision making, since people have to make decisions in a world where consequences are far from certain. Students must be familiar with the intellectual tools for modeling uncertainty and with techniques that can be used to evaluate alternatives in the presence of uncertainty.


Before entering this graduate program, students should have a good mathematical introduction to probability and statistics, including commonly used probability distributions, classical estimation, and hypothesis testing. Students will then augment their previous work in probability with a solid grounding in measure-theoretic probability, stochastic processes and simulation, as well as in related applications areas. They will apply these tools to devise better ways of modeling situations involving uncertainty, and better methods for analyzing the resulting models. Research in this field ranges from mathematical analyses of problems or solution methods to quite applied work in devising and implementing solutions to specific problems.



Organizations in industry, government, medical care, and other areas commonly encounter planning scenarios in which using scarce resources as efficiently as possible is crucial to the ability of the organization to perform successfully. In many situations, it is necessary to plan operations in a way that both explicitly evaluates many different alternatives, and rigorously accounts for constraints on resource utilization. Optimization is the area of operations research that deals with devising mathematical models and methods to identify a plan that is as good as possible in specific situations.


Students concentrating in this area should ideally have a solid introduction to calculus and linear algebra, as well as practice in modeling practical problems as mathematical programs. They will be trained to use state-of-the-art methods and software to solve these models; this training will involve instruction in the theoretical underpinnings of how these methods work, as well as significant hands-on experience in using software and interpreting its output. An important component of doctoral-level research in optimization is advancing the state of the art by devising and justifying new methods for effectively carrying out the optimization required in decision procedures. This can involve investigating either the mathematical foundations of these methods, or computational issues in designing these methods to be as effective as possible for specific kinds of problems.


Much of the research and teaching in optimization involves collaboration with the Department of Computer Sciences. More information about optimization activities at UW-Madison can be found at


Decision Analysis and Multi-Attribute Utility Theory

The decision process is goal-oriented. There are often multiple, conflicting goals to be met by a single decision. In addition, there is usually uncertainty about the consequences that will result from any given choice. In this area, students learn several kinds of techniques for structuring and facilitating such decision problems.


In particular, decision theory can be viewed as a marriage of utility theory (to express preferences about decision outcomes) and Bayesian statistics (to express uncertainty about decision outcomes). In multi-attribute utility modeling (derived from the idea of utility theory in microeconomics), students learn to construct mathematical functions to measure overall satisfaction with decision outcomes, taking into account both tradeoffs among multiple objectives and risk attitude (to account for non-linearity in preferences). Such functions can be useful to guide decision making in complex situations. In addition, Bayesian statistics and subjective probability are used to address uncertainty (even in situations where extensive statistical data are not available), and to assess the value of gathering additional information. Tools such as game theory (to address multi-player decisions), real options (to address market uncertainties), and Markov decision processes can also be useful.


Doctoral research in this area focuses on advancing the state of the art in decision analysis, either through methodological advancements, or by applying the principles of decision analysis to new classes of problems (yielding new knowledge about which solutions are likely to be optimal under particular types of conditions). A recent nationwide study of academic programs in prescriptive decision theory ( ) gave the University of Wisconsin-Madison three stars out of five (recognized for the high quality of our contributions in both education and research).



Recent dissertation titles

  • M. Bozbay (consultant at a global management-consulting firm), “Large-Scale Supply Chain Optimization via Nested Partitions”
  • M. Kilinc (post-doctoral researcher in a department of industrial engineering) “Disjunctive Cutting Planes and Algorithms for Convex Mixed Integer Nonlinear Programming.”
  • A. Krishnamurthy (faculty member in a department of decision sciences and engineering systems), “Performance Analysis of Material Control Strategies for Multi-Stage Multi-Product Manufacturing Systems”
  • S.-W. Lin (faculty member in a department of business administration), “Designing Incentive Systems for Risk-Informed Regulation”
  • M. Namazifar (senior analyst for consulting company) “Strong Relaxations and Computations for Multilinear Programming.”
  • Y. Pan (post-doctoral researcher in a department of industrial engineering), “Production Scheduling for Suppliers in Extended Manufacturing Enterprises”
  • A. Resnick (analyst in a large contract-research company), “Models for Optimizing Component Safety Stock Levels in Large-Scale Assembly Systems”
  • M. Wang (faculty member in a department of industrial engineering and management), “Guidelines for Risk-Based Inspection Programs Based on Approximate Optimal Surveillance Test Intervals”