An intelligent agent for the game of International Draughts.
The main underlying game-search mechanism is the Minimax decision rule with alpha-beta pruning. In addition, iterative deepening is exploited. This provides the algorithm with the opportunity to supply (initial) results extremely quickly.The agent has to operate in an environment with (variable) time constraints. It is, therefore, vital for the performance of the agent to employ aforementioned technique as other less scientifically sound options such as the selection of a random move result often in less sensible game-plays.
Iterative deepening also provides the opportunity to improve the search function since earlier searches can improve the alpha-beta pruning and in addition. Alpha-beta pruning is, for example, often the most efficient when it searches the best move(s) first (Artificial Intelligence: A Modern Approach (2nd ed.)).
The evaluation function that is used employs commonly used techniques and state evaluations used in the world of International Draughts. The relative importance of these heuristics is determined by an approach inspired by genetic algorithms. Instances of the agent with mutations in the values of these parameters have played in a tournament-like setting where the process of natural selection is simulated.