Hello,
I am relatively new to prtools -learning in progress:)

I’m looking for a way to achieve from classifier (mapping) not direct class label but a support function value in case, for instance, neural nets - or probabilities that given object is of class X. Generally I want to achieve soft value from given classifier. Is it easly possible or requires a lot of coding?
Is it correct assumption that after operation dataset * mapping I should get a new dataset where
resulting nlab array should contain soft labels set by mapping?

Kind regards,
Tomasz Wilk

Some classifiers give by themselves normalised soft outputs, e.g.fisherc and svc.
Other one, the ones based on density estimation need a normalisation by classc, e.g.:

a = gendath; % dataset
[trainset,testset] = gendat(a,0.5); % split in datasets for training and testing
w = qdc(trainset); % classifier
d = testset*w*classc; % compute classification matrix
+d % show confidences (here identical to posterior probs)

For classifiers that don’t need normalisation, classc is redundant, but doesn’t change anything.

Bob Duin

Thank you, that is what I was looking for :)