MAKHFI.COM
Fascinating World of Neural Nets  
Neural Network Forums
Home PageHome Page : Neural Networks for Pattern Recognition : How do I avoid overfitting?
  You are currently not logged in. You can view the forums, but cannot post messages. | HOME | Log In | Register | Search | Help
Post a Reply on This Topic Post a Reply on This Topic

Author Topic: How do I avoid overfitting?
twatson Posted: 16-Oct-03 03:57
  Edit Edit
 
Email the Author Mail   View Author's Profile Profile  
 
pejman Posted: 17-Oct-03 07:47
Delete Delete    Edit Edit
 
Email the Author Mail   View Author's Profile Profile  
use lots of training data, reduce the number of hidden units and stop the training early.
 
sammany Posted: 27-Jul-06 02:28
Delete Delete    Edit Edit
 
Email the Author Mail   View Author's Profile Profile  
You Can use Tikhonov Constant for weight elimination, Also you can use cross validation

Sammany
 
Cryxalis Posted: 12-Jul-07 08:43
Delete Delete    Edit Edit
 
Email the Author Mail   View Author's Profile Profile  
Using a single hidden layer, is preferred in most cases..

http://neurondotnet.freehostia.com/tips.html
 
punnoose Posted: 17-May-09 08:50
Delete Delete    Edit Edit
 
Email the Author Mail   View Author's Profile Profile  
early stopping, use less no of hidden neurons
 

Post a Reply on This Topic Post a Reply on This Topic
 

Copyright © 2001-2003 Pejman Makhfi. All rights Reserved.