Expected utility theory is a formal framework for decision-making under uncertainty in which a rational agent chooses the action that maximises the probability-weighted sum of the utilities of its possible outcomes. Originating with von Neumann and Morgenstern, it provides the axiomatic basis for treating preferences over uncertain outcomes as numerically comparable. It underlies Bayesian decision theory and the formulation of reward and value in Markov decision processes used throughout reinforcement learning.
Content
- Expected utility theory is a formal framework for decision-making under uncertainty in which a rational agent chooses the action that maximises the probability-weighted sum of the utilities of its possible outcomes. Originating with von Neumann and Morgenstern, it provides the axiomatic basis for treating preferences over uncertain outcomes as numerically comparable. It underlies Bayesian decision theory and the formulation of reward and value in Markov decision processes used throughout reinforcement learning.