how to fine correct classification for each class in NN

Dear All

I have 2 questions:

1- What does false/ negative, positive rate and correct neagtive, positive rate means in NN.

2- How to find total correct classification for each class.

Answer 1

[Greg Heath]

positive/negative == class 1/class 0

target output label

0        0     true negative
0        1     false positive
1        0     false negative
1        1     true  positive

The target matrix contains 0-1 column unit vectors

 trueclassindices     = vec2ind(target)
 assignedclassindices = vec2ind(output)
 err = assignedclassindices~=trueclassindices
 Nerr = sum(err)

individual class errors are ontained via

 ind1 = trueclassindices==1'

 etc

Thank you for formally accepting my answer

Greg