How to train a neural network to play games?

Question

Machine learning

Tell me how to teach neural networks to play games like checkers, chess, sea battle, crosses without zeros and other similar games?

Answer 1

Reinforcement Learning
gym.openai.com

Answer 2

• Not "like checkers", but "like checkers"
  
• You highlight the rules of the game, highlight the winning situations - and that's it, you can start a self-learning cycle. For simple games like tic-tac-toe, this can even be sorted out at random.