Advertisements
Reinforcement learning (RL) in Soar allows agents to alter behavior over time by
dynamically changing numerical indifferent preferences in procedural memory in response
to a reward signal.
This learning mechanism contrasts starkly with chunking. Whereas
chunking is a one-shot form of learning that increases agent execution performance by
summarizing sub-goal results, RL is an incremental form of learning that probabilistically
alters agent behavior.
Advertisements