Take the 2-minute tour ×
Cognitive Sciences Stack Exchange is a question and answer site for practitioners, researchers, and students in cognitive science, psychology, neuroscience, and psychiatry. It's 100% free, no registration required.

Reinforcement learning is defined not by characterizing learning methods, but by characterizing a learning problem. Any method that is well suited to solving that problem, we consider to be a reinforcement learning method.

I understand this statement. But why is it so? Why is reinforcement learning defined by a learning problem, unlike machine learning, which is characterized by a learning algorithm?

share|improve this question
Welcome to cogsci.SE. You may find that the psychological definition of reinforcement is frustratingly circular – this is a well-known criticism of the theory. Anyway, could you clarify the distinction you see between it and machine learning? Why do you think reinforcement learning is non-algorithmic? (Don't mean to imply I disagree; I'm just not sure how to apply these terms here.) –  Nick Stauner Mar 23 '14 at 18:47

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Browse other questions tagged or ask your own question.