2
$\begingroup$

For better understand of neural networks I started implementation of Multi Layer Perceptron. For now I'm implemented single Perceptron that resolve XOR problem. From this point I want start build MLP but I'm not sure if I correctly understand MLP structure. Assume I have data instance with 3 attributes and 2 classes and 5 input perceptrons. Should I give every perceptron same input and every output of perceptron should go to every hidden layer perceptron's input? I attach neural net pic. enter image description here

$\endgroup$
2
$\begingroup$

With MLPs you can follow a set of simple rules:

  1. The number of neurons in your input layer equals the number of features / data instances you have.
  2. A single neuron takes only one input (one feature)
  3. The output from one neuron can go to multiple neurons in the next layer which may or may not be fully connected.

May the force be with you ;)

| improve this answer | |
$\endgroup$

Your Answer

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

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