expectimax search
(50 minutes to learn)
Summary
Expectimax search is a search/decision-making algorithm that maximizes the average (expected) reward. It is typically applied to trees that have stochastic nodes, where the outcome of an action is uncertain.
Context
This concept has the prerequisites:
- probability (expectimax search involves stochastic (probabilistic) outcomes)
- expectation and variance (expectimax search maximizes the *expected* utility of the search)
Core resources (read/watch one of the following)
-Free-
→ EdX Artificial Intelligence
Location:
Lecture 7 Segment 2 Expectimax Search
Other notes:
- navigate between lecture material using the slider at the top
See also
- Expectimax search is an alternative to minimax search , which instead of maximizing the expected utility, minimax search minimizes the maximum loss.