Skip to content

Code structure

Dennis Trate edited this page Mar 26, 2018 · 1 revision

Code structure (?)

  1. Learning phase
    • Setup the state of the world/envrionment as an input
    • Setup the neural network in the framework to learn stuff
    • Run it a bunch of times and save the learned policy
  2. Implementation phase
    • Computer output a move (int 1-7)
    • Hardware makes the move
    • Player makes a move
    • Hardware passes player move to computer
    • Envrionment state changes based on player move
    • Computer makes a new move

Clone this wiki locally