| Interface | Description |
|---|---|
| FullModel |
An interface extension for
SampleModel that not only can provide samples of transitions, but can enumerate the probability distribution
of the transitions. |
| RewardFunction |
Defines the reward function for a task.
|
| SampleModel |
Interface for model that can be used to sample a transition from an input state for a given action and can indicate
when a state is terminal or not.
|
| TaskFactoredModel |
An interface for a
SampleModel that computes its rewards and terminal states using RewardFunction and
TerminalFunction objects that can be modified. |
| Class | Description |
|---|---|
| DelegatedModel |
An implementation of
FullModel that will delegate transition estimates for different actions to different
SampleModel or FullModel implementations. |
| FactoredModel |
A
FactoredModel is an implementation of a SampleModel or FullModel that computes transitions
using a SampleStateModel, RewardFunction and TerminalFunction. |
| FullModel.Helper | |
| TransitionProb |
Tuple for the probability of a given transition in an MDP.
|