Posts Tagged ‘Two-player games’

PostHeaderIcon Transposition Table

What is Transposition Table?

A transposition table is a table of previously encountered game states, together with their backed-up minimax values. Whenever a new state is generated, if it is stored in the transposition table, its stored value is used instead of searching the tree below the node. Transposition table can be used very effectively so that reachable search depth in chess, for example, can be doubled.

Related Articles

Two Player Games
Limited Discrepancy Search
AI Search

PostHeaderIcon Intelligent Backtracking

What is Intelligent Backtracking?

Performance of bruteforce backtracking can be improved by using a number of techniques such as variable ordering, value ordering, back jumping, and forward checking.

The order in which variables are instantiated can have a large effect on the size of the search tree. The idea of variable ordering is to order the variables form most constrained to least constrained. For example, if a variable has only a single value remaining that is consistent with the previously instantiated variable, it should be assigned that value immediately. In general, the variables should be instantiated in increasing order of the size of their remaining domains. This can either be done statically at the beginning of the search or dynamically, reordering the remaining variables each time a variable is assigned a new value.

The order in which the value of a given variable are chosen determines the order in which the tree is searched. Since it does not effect the size of the tree, it makes no difference if all solutions are to be found. If only a single solution is required, however, value ordering can decrease the time required to find a solution. In general, one should order the values from least constraining to most constraining in order to minimize the time required to find a first solution.

An important idea, originally called back jumping, is that when an impasse is reached, instead of simply undoing the last decision made, the decision that actually caused the failure should be modified. For example, consider the three-variable problem where the variables are instantiated in the order x,y,z. Assume that values have been chosen for both x and y, but that all possible values for z conflict with the value chosen for x. In chronological backtracking, the value chosen for y would be changed, and then all the possible values for z would be tested again, to no avail. A better strategy in this case is to go back to the source of the failure, and change the value of x before trying different values for y.

When a variable is assigned a value, the idea of forward checking is to check each remaining uninstantiated variable to make sure that there is at least one assignment for each of them that is consistent with the previous assignments. If not, the original variable is assigned its next value.

Related Articles

Two Player Games Search

PostHeaderIcon Limited Discrepancy Search

Limited Discrepancy Search (LDS) Algorithm

Limited Discrepancy Search (LDS) is a completely general tree-search algorithm, but is most useful in the context of constraint satisfaction problems in which the entire tree is too large to search exhaustively. In that case, we would like to search that subset of the tree that is most likely to yield a solution in the time available. Assume that we can heuristically order a binary tree so that at any node, the left branch is more likely to lead to a solution than the right branch. LDS then proceeds in a series of depth-first iterations. The first iteration explores just the left-most path in the tree. The second iteration explores those root-to-leaf paths with exactly one right branch, or discrepancy in them. In general, each iteration explores those paths with exactly k discrepancies, with k ranging from zero to the depth of the tree. The last iteration explores just the right most branch. Under certain assumptions, one can show that LDS is likely to find a solution sooner than a strict left-to-right depth-first search.

Related Articles

Two-Player Games Search Algorithms

Intelligent Backtracking

PostHeaderIcon Alpha Beta Pruning

Alpha-Beta Pruning

One of the most elegant of all AI search algorithms is alpha-beta pruning. The idea, similar to branch-and-bound, is that the minimax value of the root of a game tree can be determined without examining all the nodes at the search frontier.

Only the labeled nodes are generated by the algorithm, with the heavy black lines indicating pruning. At the square node MAX is to move, while at the circular nodes it MIN’s turn. The search proceeds depth-first to minimize the memory required, and only evaluates a node when necessary. First node and f are statically evaluated at 4 and 5, respectively, and their minimum value, 4 is backed up to their parent node d. Node h is then evaluated at 3, and hence the value of its parent node g must be less than or equal to 3, since it is the minimum of 3 and the unknown value of its right child. Thus, we level node g as <=3. The value of node c must be 4 then, because it is the maximum of 4 and a value that is less than or equal to 3. Since we have determined the minimax value of node c, we do not need to evaluate or even generate the brother of node h. Similarly, after evaluating nodes k and l at 6 and 7 respectively, the backed up value of their parent node j is 6, the minimum of these values. This tells us that the minimax value of node I must be greater than or equal to 6 since it is the maximum of 6 and the unknown value of its right child. Since the value of node b is the minimum of 4 and a value that is greater than or equal to 6, it must be 4 and hence we achieve another cut off. The right half of the tree shows an example of deep pruning. After evaluating the left half of the tree, we know that the value of the root node a is greater than or equal to 4, the minimax value of node b. Once node q is equated at 1, the value of its parent node 9 must be less than or equal to 1. Since the value of the root is greater than or equal to 2. Moreover, since the value of node m is the minimum of the value of node n and its brother, and node n has a value less than or equal to 2, the value of node m must also be less than or equal to 2. This causes the brother of node n to be pruned, since the value of the root node a is greater than or equal to 4. Thus, we computed the minimax value of the root of the tree to be 4, by generating only seven of sixteen leaf nodes in this area. Since alpha-beta pruning performs a minimax search while pruning much of the tree, its effect is to allow a deeper search with the same amount of computation. This raises the question of how much does alpha-beta improve performance. The best way to characterize the efficiency of a pruning algorithm is in terms of its effective branching factor. The effective branching factor is the dth root of the frontier nodes that must be evaluated in a search to depth d, in the limit of large d. The efficiency of alpha-beta pruning depends upon the order in which nodes are encountered at the search frontier. For any set of frontier node values, there exists same ordering of the values such that alpha-beta will not perform any cut offs at all. In that case, the effective branching factor is reduced from b to b^1/2., the square root of the brute-force branching factor. Another way of viewing the perfect ordering case is that for the same amount of computation, one can search twice as deep as with alpha-beta pruning as without since the search tree grows exponentially with depth, doubling the search horizon is a dramatic improvement. In between worst-possible ordering and perfect ordering is random ordering, which is the average case. Under random ordering of the frontier nodes, alpha-beta pruning reduces the effective branching factor approximately b3/4. This means that one can search 4/3 as deep with alpha-beta, yielding as 33% improvement in search depth. In practice, however, the effective branching factor of alpha-beta is closer to the best case of b1/2 due to node ordering. The idea of node ordering is that instead of generating the tree left to right, we can reorder the tree based on static evaluations of the interior nodes. In other words, the children of MAX nodes are expanded in decreasing order of their static values, while the children of MIN nodes are expanded in increasing order of their static values. NEXT: Quiecence

PostHeaderIcon Minimax Search

Minimax Search Algorithm

The standard algorithm for two-player perfect-information games such as chess, checkers or othello is minimax search with heuristic static evaluation. The minimax search algorithm searches forward to a fixed depth in the game tree, limited by the amount of time available per move. At this search horizon, a heuristic function is applied to the frontier nodes. In this case, a heuristic evaluation is a function that takes a board position and returns a number that indicates how favourable that position is for one player relative to the other. For example, a very simple heuristic evaluator for chess would count the total number of pieces on the board for one player, appropriately weighted by their relative strength, and subtract the weighted sum of the opponent’s places. Thus, large positive values would correspond to strange positions for one player called MAX, whereas large negative values would represent advantageous situation for the opponent called MIN.

Given the heuristic evaluations of the frontier nodes, minimax search algorithm recursively computes the values for the interior nodes in the tree according to the maximum rule. The value of a node where it is MAX’s turn to move is the maximum of the values of its children, while the value of the node where MIN is to move is the minimum of the values of its children. Thus at alternative levels of the tree, the maximum values of the children are backed up. This continues until the values of the immediate children of the current position are computed at which point one move to the child with the maximum or minimum value is made depending on whose turn it is to move.