1
$\begingroup$

I've been reading about neural networks for a long time, and I saw that in each one, the weights are always between 0 and 1. Why is this? I tried programming one, but the sigmoid function just seemed like one more thing to make it more complicated, without need. Why couldn't the range of a weight or neuron value be infinite?

Improve this question
$\endgroup$

1 Answer 1

Reset to default
1
$\begingroup$

Having the weights between 0 and 1 helps accelerate learning. They do not have to be between 0 and 1. Typically the weights get normalized to [-1, 1]. But it also depends on your problem.

Improve this answer
$\endgroup$
1
  • 5
    $\begingroup$ Can you provide some evidence for the claim "Having the weights between 0 and 1 helps accelerate learning."? Have you observed this empirically? If yes, it might be a good idea to provide the link to such an experiment. If you have seen this written in a research paper or book, cite it. $\endgroup$
    – nbro
    Mar 5 at 9:05

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .