| Interface | Description |
|---|---|
| AgentQSourceMap |
Multiagent value function planning typicall entails storing a separate Q value for each joint action for each agent.
|
| MultiAgentQSourceProvider |
An interface for an object that can providing the Q-values stored for each agent in a problem.
|
| QSourceForSingleAgent |
And interface for objects that can store and retrive Q-value objects for a specific agent and state and joint action query.
|
| SGBackupOperator |
A stochastic games backup operator to be used in multi-agent Q-learning or value function planning.
|
| Class | Description |
|---|---|
| AgentQSourceMap.HashMapAgentQSourceMap |
An implementation of the
AgentQSourceMap in which the sources are specified by a hash map. |
| AgentQSourceMap.MAQLControlledQSourceMap |
An implementation of the
AgentQSourceMap in which different agent objects each maintain their own personal Q-source. |
| JAQValue |
Class for storing Q-value informartion for a joint action.
|
| MAQSourcePolicy |
An abstract extension of the JointPolicy class that adds a required interface of being able to a
MultiAgentQSourceProvider. |
| MAValueFunctionPlanner |
An abstract value function based planning algorithm base for sequential stochastic games that require the computation of Q-values for each agent for each joint action.
|
| QSourceForSingleAgent.HashBackedQSource |
An implementation of the
QSourceForSingleAgent interface that has Q-values backed by hashmaps. |