In a team of three students, we used various machine learning algorithms to create a program, which plays checkers without using any checkers library. We used such algorithms like minimax and alpha-beta pruning, but also neural networks with backpropagation and eventually we got to compare them and choose the one that wins the most rounds (against itself, but also against a human).