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.

Register Log in

[SOLVED] having trouble with neural networks

Status
Not open for further replies.
S

sadatwins

Junior Member level 2
Junior Member level 2
Joined
Apr 24, 2010
Messages
20
Helped
1
Reputation
2
Reaction score
1
Trophy points
1,283
Location
jordan
Activity points
1,414
can you please help me.:-o
my problem is that every time i run my code performance varies (i mean accuracy of classification change)i'm using neural network for P300 Classification.
i train my neural network with 3 sessions and i use 4th sessions to test neural network.

here is my matlab code

P=x2;
Tg=y;
net=newff(P,Tg,5 );
[net,tr]=train(net,P,Tg);

and to test my data:
y=sim(net,x2);

here is performance figures(i got them by comparing how many of test data were classified correctly):
untitled.png
Capture.PNG
 

Hi, it is the normal the accuracy of recognition using NNs will change for every execution, because NNs use random weight adjustment. So u need to take may be several results and average it.
 
How many classes do you have? I believe if you increase the no of inputs then you will get better results.
 
Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top