Reinforcement Learning

the process of training an agent to perform a series of tasks towards a greater goal can take many avenues. reinforcement learning is the process of using a series of generated actions with respect to the agent environment that generates the best long-term outcome.

agents pull actions and the resulting reward from completing the action provides the main learning level for this algorithm; rewards determine which actions will be done in a similar state in the future and can be tuned adaptively, to be respective to the field of options available. the goal is signified through a policy that encompasses the agents decision making core.


book: Reinforcement Learning - An Introduction