Quantcast
Channel: MATLAB Central Newsreader Recent Posts
Viewing all articles
Browse latest Browse all 19628

Re: neural nets - trainbr & regularization penalty

$
0
0
Yes, I have seen the doc.

So it sounds like it is not possible for a user to choose the value for the regularization parameter, and the documentation doesn't say how it chooses the value, and I'm guessing it isn't possible to even see what value the function chose either.

This is a poor situation for users to be in, regarding a pretty basic option for network training.

Viewing all articles
Browse latest Browse all 19628

Trending Articles