SKEDSOFT

Description:-The way of playing games by revealing consequences through applying logic and heuristic rules is only one approach to game playing. Another approach, mainly applicable to simple board games, is to learn how to react to certain patterns on the board.

Such a stimulus-reaction approach can be represented by associations in the form of:

IF pattern i , THEN move j

This strategy does not require the program (the computer system) to know many concepts, like rows, columns, etc. The program can learn how to react to certain patterns. This is exactly what neural networks can do. This approach is implemented by setting properneural network architecture and training the network with associations of the type:

input pattern → output move

That is, it is trained with "good" moves; "bad examples" or examples of "what not to do" should not necessarily be used in this approach as they may be too many.

Example: Playing Ticktacktoe by MLP, afeed forward neural network for training with good examples for playing ticktacktoe. A disadvantage to this approach is that for bigger game boards and games that are not as simple, many examples of good moves are required, which may be impossible to supply.