Download Latest Version tictactoenn.zip (25.5 kB)
Email in envelope

Get an email when there's a new version of 5_5_5_tic_tac_toe_NN

Home / resources / trainingExemplars
Name Modified Size InfoDownloads / Week
Parent folder
opening_moves_O_before.txt 2011-08-09 60 Bytes
opening_moves_O_desired.txt 2011-08-09 253 Bytes
OffensiveX.txt 2011-08-09 687 Bytes
OffensiveXDesired.txt 2011-08-09 2.6 kB
OffensiveODesired.txt 2011-08-09 2.7 kB
DefensiveXDesired.txt 2011-08-09 2.1 kB
OffensiveO.txt 2011-08-09 706 Bytes
DefensiveODesired.txt 2011-08-09 2.0 kB
DefensiveX.txt 2011-08-09 842 Bytes
DefensiveO.txt 2011-08-09 813 Bytes
correct_patternsX.txt 2011-08-09 4.9 kB
correct_patternsO.txt 2011-08-09 4.9 kB
before_correct_patterns.txt 2011-08-09 8.1 kB
Totals: 13 Items   30.6 kB 0
This file contains background information on the tictactoenn repository.  The files contained within the repository are: (1) code for an artificial neural network that uses error backpropagation with the intent of playing a 5x5x5 game of tic-tac-toe, (2) a training program for the artificial neural network, (3) training exemplars for the artificial neural network, (4) code of an opponent making random moves, (5) a datafile for the neural network's weights, and (6) code for the 5x5x5 tic-tac-toe game implementation.  Shahzeb Siddiqui and Francis Mutuc are the creators of the artificial neural network agent used to play this 5x5x5 game of tic-tac-toe. Any usage of any of this code should be given proper credit to its creators.
Source: README.txt, updated 2011-08-09