Generalising to unknown models
The ability to plan is an important part of human intelligence, allowing us to solve problems and make decisions about the future. For example, if we see dark clouds forming, we might predict it will rain and decide to take an umbrella with us before we venture out. Humans learn this ability quickly and can generalise to new scenarios, a trait we would also like our algorithms to have.
Researchers have tried to tackle this major challenge in AI by using two main approaches: lookahead search or model-based planning.
Systems that use lookahead search, such as AlphaZero, have achieved remarkable success in classic games such as checkers, chess and poker, but rely on being given knowledge of their environment’s dynamics, such as the rules of the game or an accurate simulator. This makes it difficult to apply them to messy real world problems, which are typically complex and hard to distill into simple rules.
Model-based systems aim to address this issue by learning an accurate model of an environment’s dynamics, and then using it to plan. However, the complexity of modelling every aspect of an environment has meant these algorithms are unable to compete in visually rich domains, such as Atari. Until now, the best results on Atari are from model-free systems, such as DQN, R2D2 and Agent57. As the name suggests, model-free algorithms do not use a learned model and instead estimate what is the best action to take next.
MuZero uses a different approach to overcome the limitations of previous approaches. Instead of trying to model the entire environment, MuZero just models aspects that are important to the agent’s decision-making process. After all, knowing an umbrella will keep you dry is more useful to know than modelling the pattern of raindrops in the air.
Specifically, MuZero models three elements of the environment that are critical to planning:
- The value: how good is the current position?
- The policy: which action is the best to take?
- The reward: how good was the last action?
These are all learned using a deep neural network and are all that is needed for MuZero to understand what happens when it takes a certain action and to plan accordingly.