Topic outline

  • Topic of the seminar: Decision making under uncertainty
    Teachers: Antti Punkka (, Jussi Leppinen


    The seminar offers an introduction to models and methods for decision making under uncertainty. More specifically, discussed topics include

    • modeling uncertainties with probabilities, ranging from simple markov chains to Bayesian updating and Bayesian networks,
    • decision trees, influence diagrams and utility theory as means for modeling and solving decision making problems with uncertain outcomes,
    • dynamic programming, with emphasis on markov decision processes and their variants to model and solve optimal policies over finite and infinite time horizons.

    Theoretical topics will be complemented by presentations on reported applications.

    The exact topics to be covered depends on the number of students and will hence be announced later.

    Practical matters

    Seminar meetings: on Fridays 9:15 - 12:00 (Y228a). First meeting on Friday Sep 18!

    Assessment methods: Home assignments and presentations; grading principles:

    • Maximum points 100. Grading principles:
    1. Presentations max 60 points - minimum requirement: both presentations 'passed'
    2. Home assignments max 30 points - minimum requirement: 15 points
    3. Participation (attendance) in seminar meetings: max 6 points - minimum requirement: ~80% = 27 h (/ 33h)
    4. Successful submission of course feedback: 4 points

    Grading scale: 0-5

    Study material: TBA

    Language of instruction: English

    Prerequisites:  MS-C2105 Introduction to Optimization (or equivalent); MS-E2134 Decision Making and Problem Solving and MS-C2111 Stochastic Processes are useful