Module rlr::agents

Agents facilitate the learning process for a model.

Specifically, they provide methods for the following:

• Recommending an action given a state.
• Transitioning from one state to another state given some action.
• Learning from the level of success achieved when moving from one state to another via some action.

Modules

bayesian

Contains an agent that weights the initial condition of each potential action based on the number of observations of that action relative to the number of observations for all other actions.

Traits

Agenter

Represents something that is capabile of recommending actions, applying actions to a given state, and learning based on the transition from one state to another.