Multi-Layer Perceptron (MLP), Deep AutoEncoder (DAE), and Deep Belief Network (DBN) are trained differently.

However, do they follow the same process during the inference phase, i.e., do they calculate a weighted sum, then apply a non-linear activation function, for each layer until the last layer, or is there any difference? Moreover, are they only composed of fully connected layers?

  • 1
    $\begingroup$ Could you please focus on the comparison of 2 models at a time rather than 3? Moreover, when you use an acronym, like DAE, you should first use the long/full version, like "muli-layer perceptron (MLP)", so that everyone knows what the acronym refers to. $\endgroup$
    – nbro
    Apr 5 '21 at 2:18
  • $\begingroup$ I kept the three because the goal is not to know the detailed differences between the models .. but just to know if they are all based on "only" fully connected layers and thus they perform the same mathematical operations during inference phase. $\endgroup$
    – witdev
    Apr 5 '21 at 17:45
  • 1
    $\begingroup$ I made some changes to your post in an attempt to improve the clarity. Please, check that I didn't change the meaning of the post. $\endgroup$
    – nbro
    Apr 9 '21 at 3:17

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.