Interprets the essence - how best to get down to this business. I have 2 ideas in my head:

  1. A clear set of instructions in the form of a tree
  2. Self-learning in the process of playing the game "you are like that - it means that I’m getting worse - a bad move, I won’t do it anymore, it turns out better - a good one, you have to remember and use it."

Tell me if someone solved similar problems.

    3 answers 3

    There is an approach to using neural networks that train on a sequence (LSTM, RNN, etc.) for strategic games. In this case, the "sequence" means a series of moves by players. Example: A Scalable Neural Network Architecture for Board Games . In general, artificial intelligence has a lot of good stuff here: Artificial Intelligence .

      Try state machines

      In general, there is a good book on this topic.

      • In the book, more and more about FPS, like, not? But it doesn’t matter, of course - neoascetic

      Udacity has a great AI course . True, there is not for games, but in general.