- Just for fun
- I'm in the process of learning C++
- Using SFML as the GUI component (still unpolished)
- Educational
- (Failing) TD-Leaf λ temporal difference learning - in progress
- Adversarial search (Minimax with alpha-beta pruning, Iterative deepening minimax, etc ...)
The board has the following layout:
(Player V's goal)
+----------+
|H |
|H | (player H's goal)
|H |
| V V V |
+----------+
The aim of the game is simple - given a NxN board, move all your pieces across the opposite side of the board.
The rules are:
H
is allowed to move North, South, and EastV
is allowed to move North, West, and East
(i.e. pieces are not allowed to move "backwards" or diagonally)
note that most of the learning algorithms are not working as expected and the logic needs to be reworked.
more instructions coming soon on compilation and running the GUI module.