Bayesian Decision Analysis

The course provides formal methods and models intended to help decision makers facing uncertainty, to consistently analyze, model and resolve their choice problems.

We adopt a formal approach with an emphasis on understanding how to model and measure decision makers’ beliefs (regarding uncertainties) and preferences (regarding monetary and non-monetary outcomes). Beliefs and preferences are analyzed and measured using techniques based on (i) Bayesian inference and reasoning and (ii) Rational choice principles from modern decision theory, including decision making in the presence of multiple objectives.

Course Objectives

The course has the following specific objectives:

  • Learn Bayesian techniques of inference and reasoning (including how to build Bayes nets using a specialized software)
  • Measure subjective beliefs (probabilities) under uncertainty (the course provides techniques to build subjective probability distributions)
  • Learn how to model decision problems using Bayesian Networks, Influence Diagrams and Decision Trees
  • Decision making with multiple objectives.

Course Content

The course consists of four parts:

  1. Bayesian Inference and Reasoning
  2. Measuring Beliefs Under Uncertainty
  3. Modeling Decision Problems
  4. Multi-objective Decision Making (modeling and measurement)