What people are saying - Write a review
We haven't found any reviews in the usual places.
Search and Learning
The characteristics of the consistency search for a difficult win
6 other sections not shown
alpha-beta analysis Artificial Intelligence board games branching factor brute force search capture tree Chapter children of node coerced node connect-four consistency search procedure consistent or coerced current board position depth determine dissertation evaluation function error evaluation function value evaluation value Figure III.3 Figures V.4 fixed-depth full-width game of chess game tree game-learning program games of tic-tac-toe games played given GNUCHESS height h incorrect move incorrectly evaluated independent such searches learning algorithm legal moves let node LOSE status machine minimax negamax value node h non-coerced node number of games number of moves one-ply search perfect information piece types player whose turn playing the game ply search problem q-level raw board position root node root position SAL program SAL versus SAL's performance search is performed search tree search value selective search shown in Figure Temporal Difference Learning two-ply search type of piece value of node value of position WIN status won position