i am confused with RBF kernal based ANN classification. when i implemented, all images are missclasified.. T is class label , P is training image feature , P1 is testing image feature...Can any one correct my code please,...​..........​..........​.

%%%%%%%% RBF training %%%%%%%%%%%
Tc=[1 1 1 1  2 2 2 2 2 2 2 2 2 ];
SPREAD=1;
T=ind2vec(Tc);
net=newrbe(P,T,SPREAD);
%%%%% test data %%%%%%%%
P1= FF1;
Y= sim(net,P1);
ANNresult = vec2ind(Y);

Answer 1

[Greg Heath]

P and FF1 are undefined

>> net = newrbe(P,T,SPREAD);

Undefined function or variable 'P'.

>> P1= FF1;

Undefined function or variable 'FF1'.

Hope this helps.

Thank you for formally accepting my answer

Greg

[sruthi vs]

Sir i already specified variable P and P1 in my program..P store training feature value ...P1 stored testing features.Sir I want code for how test the image...Sir please help me to correct implementation. In feature extraction step i stored features for training and testing in P and P1

[Greg Heath]

If P and P1 are fixed, there is only one way to improve your performance:

 Vary SPREAD

Otherwise, redivide P+P1 and try again.

Hope this helps.

Greg

[sruthi vs]

again missclassified value obtained .... I input 13 test image.5 noncancer class and 9 cancer class..but only first image show non cancer and all remaining image show cancer at all time....it is a wrong classification sir....What i will do sirrr?????