DeeperMind

Best M(AI) group

create environment

$pip3 install -r requirements.txt
$python3 game.py {number_rows} {number_columns} {connect_number}

How to

1. Initialise Game object.
2. Use functions, go wild.
3. You need to create a loop, like in main.py because the game is turn based.

Only function you will need to access from "Game" object I think.

Checking functions

• is valid location
• get valid moves
• check for win
• get next open row

Interaction functions

• drop piece
• next turn

Display functions

• draw board (pygame)
• print board (terminal window)

Algos

• DFS & BFS
• Minimax with alpha beta pruning
• Q Learning

Q Learning

python q_learning.py {Training Mode} {Iterations}

i.e python q_learning.py 1 5 To train python q_learning.py 0 5 To play

To play, set Training Mode to either 1 or 0. 1 = Training Mode ON 0 = Training Mode OFF

Q-learning uses its own Minimax file, minimax_agent_for_q_learning.py. This is because the q-learning agent must be the first one to start (because of the way we codede it)