人工智能的课件CH6_Game.ppt

* Maxn Pruning Why is maxn weak in practice? Only compares 2 scores out of n players Relies on game evaluation properties, not ordering Last-Branch Pruning Speculative Pruning Last-Branch/Speculative Pruning 1 (3, 3, 4) (3, 3, 4) 2 (3 Players) 2 3 3 (1, 4, 5) 1 (2, 4, 4) 2 Last Branch/Spec. Pruning Best case: O(bd·(n-1)/n) As b gets large Dependent only on node ordering in tree /~nathanst/ for more info Chance Dices, Cards, Scrabble, etc Chance in games started as a “divine” mechanism: Relation between divination and games, e.g. Tarot Many “road of life” games (Ganzebord, Snakes and ladders, Parchisi, Seth) Chess started as a racing game with dice Chance is introduced in the game tree by chance-nodes (a third player) Equilibria are not pure anymore (the best way to move might be a mixture of two moves) Chance Strategy 1: Use *-minimax by Ballard 1983 At min nodes minimize, at max nodes maximize, at chance nodes determine the expected payoff Might not produce optimal strategy Works with dice, not with card decks or scrabble Strategy 2: use Monte-Carlo Simulation Generate a random configuration of cards Compute an optimal strategy (Alphabeta) Repeat many times Determine move with best average pay-off Used in Scrabble, Bridge, Poker, etc Games of Chance Backgammon is a two player game with uncertainty. Players roll dice to determine what moves to make. White has just rolled 5 and 6 and had four legal moves: 5-10, 5-11 5-11, 19-24 5-10, 10-16 5-11, 11-16 Such games are good for exploring decision making in adversarial problems involving skill and luck. Game Trees with Chance Nodes Chance nodes (shown as circles) represent the dice rolls. Each chance node has 21 distinct children with a probability associated with each. We can use minimax to compute the values for the MAX and MIN nodes. Use expected values for chance nodes. For chance nodes over a max node, as in C, we compute: epectimax(C) = Sumi(P(di) * maxvalue(i)) For chance nodes over a min node compu