SYST 473 / 573 / OR 681 : Decision and Risk Analysis
Thursday, , Science & Technology II, room 9
Texts and Reading Materials:
(1) Clemen and Reilly, Making Hard Decisions, 2nd ed., 2001. In bookstore.
(2) Decision Tools software provided with book (requires Excel)
(3) Class notes. I will generally post a set of PowerPoint slides on the web site when I use the slides in the lecture. I try to post them before class, but sometimes I revise them at the last minute, in which case they will be posted after class.
(4) Class handouts. Posted to web page before class.
Class Web Page: http://classweb.gmu.edu/myoungre/SYST573
This course is a study of analytic techniques for rational decision making that addresses uncertainty, conflicting objectives, and risk attitudes. This course covers modeling uncertainty; rational decision making principles; decisionmaking under uncertainty without use of probability; representing decision problems with utility functions, decision trees, and influence diagrams; solving multi-attribute utility functions, decision trees and influence diagrams; defining and calculating the value of data and information; incorporating risk attitudes into the analysis; and conducting sensitivity analyses.
Course Objectives: Students will learn the concepts associated with rational decision making and will learn how to structure and solve simple decision problems associated with conducting systems engineering trade-offs and managing systems engineering projects.
Assignments and Grading:
· Readings: Readings will be assigned weekly; students are expected to complete readings before class.
· Homework: Homework will be assigned but not graded. Solutions will be posted and discussed during the next class period after assignment. Although not graded, students are highly encouraged to complete the homework, as the exam questions will be similar to the homework assigned.
· Undergraduate students (SYST 473): There will be a midterm and a final exam. The midterm and final exam will constitute 40% and 60% of the grade respectively.
· Graduate Students: There will be a midterm, a final exam, and a project. The midterm, final exams and the project will constitute 35%, 45% and 20% of the grade respectively.
Graduate Project: The project can be done individually or as part of 2-3 person group. The project should address a real life decision with either a personal or business focus. It will work best if a friend or colleague was facing the decision, but usually students pick a decision of their own. Occasionally students select a decision that was made in the past and perform a retrospective analysis; this approach is difficult because there is a lot of research that must be done to capture the situation realistically. Students shall submit a statement about whether they are going to do an individual or group project by the end of third class period. An abstract of the decision that is the focus of the project shall be submitted by the end of the sixth class period. The abstract should contain a summary of the decision and lists of tentative objectives and alternatives.
Project Requirements for Decision/Risk Analysis
1. Provide a summary of the decision problem (who is the decision-maker, what is the decision time frame).
2. Define the objectives hierarchy and utility functions (what are the objectives and why are they objectives, what is the utility function for each bottom level objective and what were the assessment queries used to define these functions, what are the weights for the objectives
hierarchy and what were the assessment queries used to define these weights).
3. Define the alternatives.
4. Define any uncertainties included in the analysis (define the variables, the uncertain states, the conditioning variables, and the probability distributions).
5. Summarize the results of the analysis. Use decision trees and influence diagrams (if appropriate).
6. Describe the analysis computations.
7. Provide sensitivity analyses for parameters that might change the recommended decision.
8. More detail will be provided by the sixth period.
Overview & Discussion of Values, Measures, and Decisions
Simple Decisionmaking Using Non-Stochastic and Stochastic Criteria
Decision Structure: Influence Diagrams and Decision Trees
Making Decisions: Solving Influence Diagrams and Decision Trees
Modeling Uncertainty: Probability Applied to Decision Analysis and Risk Management
Risk Attitudes and Utility
Multi-Attribute Utility Functions
Advanced Topics in Decision Analysis (if time permits – not graded)
* Note: A topic may last more or less than one lecture period each depending on the topic complexity and class discussion
Schedule: Class every Thursday, from January 23rd to May 1st