Punishment is reducing a behaviour due to bad outcome e.g. A cow stops touching an electric fence because it gets a shock.

Negative reinforcement is increasing a behaviour that reduces a bad outcome. e.g. Put on a seat belt to stop the car bleating.

In this context, how are neural nets trained by Reinforcement Learning (RL) to approximate pi(s) and Q(s,a) ? Must negative neuron activations be allowed in hidden layers to enable negative reinforcement ? Are there practical issues with value ranges ?

  • 1
    $\begingroup$ What research have you done? Where have you looked? What books have you read? There are many standard resources that describe how reinforcement learning works, and many of them describe how to use neural networks with RL. There seems little point in us repeating that material. Instead, I suggest that you spend some time reading about RL, then if you're still unsure about something specific, tell us what you read and what you remain confused about. $\endgroup$ – D.W. Jun 14 '17 at 15:59
  • 1
    $\begingroup$ P.S. I don't think that trying to understand the English-language definition of terms like "punishment" and "negative reinforcement" is going to help you understand how RL works in detail. That phrase is a loose metaphor, not a precise description. $\endgroup$ – D.W. Jun 14 '17 at 15:59

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.