Keras Reinforcement Learning Projects
by Giuseppe Ciaburro, Sudharsan Ravichandiran, Suriyadeepan Ramamoorthy
DeepMind AlphaZero
AlphaZero is an artificial-intelligence algorithm based on machine-learning techniques developed by Google DeepMind. It is a generalization of AlphaGo Zero, a predecessor developed specifically for the game of Go, and in turn an evolution of AlphaGo, the first software capable of achieving superhuman performances in the game of Go. Like AlphaGo Zero, it uses the Monte Carlo Tree Search (MCTS), guided by a deep convolutional neural network trained for reinforcement.
On December 5th 2017, the DeepMind team published a preprint on arXiv in which some of the results obtained by AlphaZero were presented in several classic table games, reaching a superhuman level in the game of chess, shogi, and Go with only a few hours of training, ...
Become an O’Reilly member and get unlimited access to this title plus top books and audiobooks from O’Reilly and nearly 200 top publishers, thousands of courses curated by job role, 150+ live events each month,
and much more.
Read now
Unlock full access