The document discusses algorithms used for game playing in artificial intelligence, including the minimax algorithm and alpha-beta pruning algorithm. It provides an overview and examples of how the minimax algorithm works using a game tree to evaluate all possible moves to minimize losses and maximize wins. It then explains how the alpha-beta pruning algorithm improves upon minimax by pruning branches of the game tree once it determines the optimal solution.
1. Topic :
Game playing in AI
• PRESENTED BY
• VICKY TYAGI
• MCA 3RD SEM
• 2001560001
2. Overview of presentation
• Algorithm in game playing
• Popular algorithm in game playing
• Minimax algorithm
• Alpha beta pruning algorithm
• Most suitable algorithm
3. Algorithm in game playing
• Any game works on a particular algorithm
• Most strategy based game uses minimax algorithm or alpha-
beta pruning algorithm
• Games like chess or tic-tac-toe uses these algorithm
• The algorithm works as every action possibly took by player and
then compare the action with the computer move and then take
action
• These games uses particular set of action and follow particular
set of rules
4. Minimax algorithm
• Here, there are two things i.e minimum and maximum value
• Maximum is a move done by user that aims at maximizing the
chance of winning
• Minimize is a move done by AI that aims at minimizing the
chance of losing
• Let’s have a example for better understanding
5. Example for Minimax algorithm
• As shown in fig 1 given the algorithm have tree to traverse from
root A to the terminal nodes
• The algorithm assumes max as player(human) and min as
AI(machine)
• The algorithm is much similar to DFS as the action is taken after
traversing last node comparison.
7. Advantages and disadvantages
• The minimax algorithm is not suitable for large game tree.
• The minimax algorithm is easy to implement
• The time complexity of minimax algorithm is O(b^d).
• Every move in game by AI(computer) is taken after several
comparisons.
8. Alpha Beta pruning algorithm
• As soon as the optimum solution is found the algorithm
immediately neglects any other solution possible.
• Generally alpha is considered for max node and beta is
considered for min node.
• The aim of algorithm is to have solution in minimum number of
moves
9. Fig 2: A game tree for Alpha Beta Pruning algorithm