muzero - XoilacTV

MuZero is a new algorithm that combines tree-based search with a learned model to master various complex domains without any knowledge of their dynamics. It achieves superhuman performance on 57 Atari games and matches AlphaZero on Go, chess and shogi.

DeepMind's MuZero was created in order to overcome limitations by the two previous planning methods. In order to do that, MuZero focuses on modeling only the aspects of the environment that are relevant for the algorithm to make its decisions.

MuZero is a method that combines a learned model with search to plan in various tasks, such as Atari, Go, Chess and Shogi. The model predicts policy, value and reward from hidden states that are updated by a recurrent process.

MuZero is a new approach to model-based reinforcement learning that combines a tree-based search with a learned model of the environment's dynamics. It achieves superhuman performance in Atari games and matches AlphaZero in chess, shogi and Go without any knowledge of the rules.