This is the Django-Backend that holds the game-ai logic along with the neural net framework.
Contributors- Eric Enderlein, Chris Hudson, Alex Short, W. Ely Paysinger, Sean Beseler
About- Neural.py has holds the class framework necessary to make a neural network instance. This is a blank slate neural net that is in essence modular by design. Genetic.py calls an inherited Neural class as Network, modifies the class to have tic-tac-toe specific methods. Along with Network, Genetic also has Game, Generation, and Individual as classes. Game translates board states from the front-end into data readable for Neural. Generation holds methods for exports and import, generating test boards, selecting a random individual neural network within a generation, and several run methods, which create different fitness programs for neural networks. Individual holds all of the mutation/reproduction code for neural networks. Individual also has evaluate and class attributes have unique neural network stats.
Reference code: https://github.com/llSourcell/tic_tac_toe_neural_network-LIVE