I've always been a bit confused when it comes to Deep Learning terminology.

Is the definition of the perceptron, whether single layer or multi layer, associated with a specific type of activation function? (aka Step function?)

Or is it associated with any activation function that attempts to squash the output in between a range of values? (aka sigmoid, TanH..)

The definition of a perceptron is still not very clear to me, despite all the sources i've read.


1 Answer 1


The term multilayer perceptron is used as a synonym for a normal feedforward neural network. It doesn't necessarily imply a specific kind of activation function.

  • $\begingroup$ Well according to the wiki, the perceptron is an algorithm, under Supervised Learning, used to handle binary classifier-based problems. So if that is the case, then the activation function would have to be geared towards binary classifiers in general hinting at specific activation functions $\endgroup$ Commented Oct 20, 2020 at 9:37
  • $\begingroup$ Yes, that's a sensible thing to assume. However that's not how the term is used today, which makes it rather misleading. I think it's better to avoid using the term. $\endgroup$
    – Tom Dörr
    Commented Oct 20, 2020 at 11:59

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.