Jump to content
  • Advertisement

Archived

This topic is now archived and is closed to further replies.

Tac-Tics

The bias

This topic is 5986 days old which is more than the 365 day threshold we allow for new replies. Please post a new topic.

If you intended to correct an error in the post then please contact us.

Recommended Posts

I''ve been reading about NN''s and it''s all very interesting stuff. I have a question though about the bias for a NN. Does each individual Neurode have a bias or is one bias assigned to every Neural Layer? I''m just asking because I read two different tutorials that showed two slightly different representations and they didn''t really mention on what level the bias exists. Any help = thnx "I''''m not evil... I just use the power of good in evil ways."

Share this post


Link to post
Share on other sites
Advertisement
well normally every neuron has a bias. But you can also use no biases, but connect every neuron via a weighted connection to a always full active neuron and then you don''t have to have an explicit learning rule for biases, you only have to update weigths. visualizations of this nets with so called on-neurons look a bit crappy, cause there are this lot of connections to this one neuron, but well, this can be solved nicely, too at the moment i cannot think of other disadvantages ...

@$3.1415rin

Share this post


Link to post
Share on other sites

  • Advertisement
×

Important Information

By using GameDev.net, you agree to our community Guidelines, Terms of Use, and Privacy Policy.

We are the game development community.

Whether you are an indie, hobbyist, AAA developer, or just trying to learn, GameDev.net is the place for you to learn, share, and connect with the games industry. Learn more About Us or sign up!

Sign me up!