Please enable JavaScript.
Coggle requires JavaScript to display documents.
Adversarial search - Coggle Diagram
Adversarial search
Zero-Sum Games
in game theory, a zero-sum situation is one in which the gains of one player are equal to the losses of another, resulting in no net change in wealth or benefits.
AI focuses on games with abstract, well-defined rules and states.
-
Utility values are equal and opposite (I win, you lose, or we draw).
Strategy:
Partially Observable & Non-Deterministic: Agent relies on perceptions to make decisions. (e.g., Wumpus World). Requires contingency plans.
Fully Observable & Deterministic: Agent knows action outcomes, can calculate next state. (e.g., map navigation)
Solving Games
-
Challenges: Exponential growth of search tree (e.g., chess).
-
-
Alpha-Beta Pruning
Optimization:
Prunes away parts of the search tree that cannot influence the optimal decision, significantly reducing computation.
-
-
what is Adversarial search
An adversarial search is one in which we investigate the issue that emerges when we attempt to prepare ahead of time while other agents are preparing an attack against the United States.