Course Outline
This course
consists of two parts: fundamentals and applications.
Fundamentals
This part of the
course consists of lectures given by the instructor.
Introduction:
motivation, representing uncertainty
Elementary
decision theory:
- Utility and descriptive statistics
- Uncertainty due to ignorance of the state of the nature
- Bayes strategies
- Estimation
- Minimax decision rules and games
- (Briefly) probabilistic networks and graphical models
Multi-stage
decision making:
- MDPs and POMDPs
- Dynamic Games
Applications
Within three
weeks from the first lecture, each student will pick a problem where
dealing with uncertainty plays a fundamental role and identify
solutions proposed in the literature. Ideally, there will be 2-3
seminal papers addressing the same problem using different approaches.
The second part
of the course will consists of lectures given by the students. Each
student will give a lecture which
- introduces and motivates the chosen problem,
- explains proposed methods in the literature (a high-level
presentation is highly
undesirable)
- compares and critiques these methods.
At the end of
the semester, students are required to submit a 3-4 page term-report on
their chosen problem.
Grading
- Take-home exam: 40%
- Presentation 25%
- Term-report: 20%
- Class participation: 15%
In grading the
presentation and the term-report, the following will be important:
understanding of the fundamentals and comprehensiveness
of the critique.