Please enable JavaScript.

Coggle requires JavaScript to display documents.

Adversarial Search, Uncertainty in outcomes (like Wumpus World), Example…

- - - - To identify the best strategy considering possible moves of an opponent.
    - - Chess, tic-tac-toe players compete directly, each affecting the other's success.
- - - - Min: Aims to maximize chances of winning
    - - Terminal-Test: Checks if the game is over.
        Utility Function: Assigns numerical values (win = positive, lose = negative).
    - - Agent (Max) tries moves to win, assuming opponent (Min) tries to block it.
- - - - Speeds up the search by not considering ineffective moves.
        
        Result: Dramatically reduces time and complexity of decision-making.
        
        Alpha (α): Highest score Max can guarantee.
        Beta (β): lowest score Min can guarantee.
- - - - Gradually searches deeper into game trees
    - - Stores moves that previously resulted in favorable outcomes
    - - Uses memory to remember previously visited states, avoiding redundant checks.
- - - - Decisions must account for uncertainty
  - - - Monte Carlo Roll-out: Uses repeated random simulations to guide decisions
- - - - Shows the conflict between short-term self-interest and long-term collaboration