A set of cognitive neuroscience inspired agents and learning algorithms.
These consist of implementations of the canonical Q-Learning, Actor-Critic, Value-Iteration, and Successor Representation algorithms.
All algorithms are "tabular" and work with observations that consist of integer representations of the state of the agent. This corresponds to the index observation type.
The implementations of the TD algorithms can be found here.
- TD-Q
- TD-SR
- TD-AC
The implementations of the Dyna algorithms can be found here.
- Dyna-Q
- Dyna-SR
- Dyna-AC
The implementations of the model-based algorithms can be found here.
- Value Iteration (MBV)
- TDSR / Value Iteration Hybrid (SRMB)