Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

neural network training

Status
Not open for further replies.

Demonis

Member level 1
Joined
Jun 1, 2006
Messages
36
Helped
1
Reputation
2
Reaction score
0
Trophy points
1,286
Activity points
1,528
What does it mean when after a few epochs the training is stopped with "Minimum gradient reached". It happens with neural network with 2 or 3 hidden layers (20-30 neurones in a layer). I make a training in Matlab with trainbr-function. After such a short training ANN is not good at all (a lot of errrors more than 50%).
 

You should try various of parameter in the neural network training. I never use the built in neural network in matlab, in fact, I write the neural network myself.
 

The analysis with SOM NN showed that different input data produces approximately the same output value . Perhaps that the reason for poor training???
 

Im not very sure about the real reason, as I have no idea what type of NN you're using.

For SOM, this NN is suitable for categorization and grouping.
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top