FDSC

Trainable Feature based Dissimilarity Space Classifier

    W = FDSC(A,R,FEATMAP,TYPE,P,CLASSF)
    W = A*FDSC([],R,FEATMAP,TYPE,P,CLASSF)
    D = X*W

Input
A	Dateset used for training
R	Dataset used for representation  or a fraction of A to be used for this.  Default: R = A.
FEATMAP	Preprocessing in feature space (e.g. SCALEM) Default: no preprocessing.
TYPE	Dissimilarity rule, see PROXM Default 'DISTANCE'.
P	Parameter for dissimilarity rule, see PROXM Default P = 1.
CLASSF	Classifier used in dissimilarity space, default LOGMLC
X	Test set

Output
W	Resulting, trained feature space classifier
D	Classification matrix

Description

This routine builds a classifier in feature space based on a  dissimilarity representation defined by the representation set R and the dissimilarities found by A*FEATMAP*PROXM(R*FEATMAP,TYPE,P).  FEATMAP is a preprocessing in feature space, e.g. scaling  (SCALEM([],'variance') or pre-whitening (KLMS).

R can either be explicitely given, or by a fraction of A. In the  latter case the part of A that is randomly generated to create the  representation set R is excluded from the training set.

New objects in feature space can be classified by D = B*W or by  D = PRMAP(B,W). Labels can be found by LAB = D*LABELD or LAB = LABELD(D).

Example(s)

 a = gendatb([100 100]);    % training set of 200 objects
 r = gendatb([10 10]);      % representation set of 20 objects
 w = fdsc(a,r);             % compute classifier
 scatterd(a);               % scatterplot of trainingset
 hold on; scatterd(r,'ko'); % add representation set to scatterplot
 plotc(w);                  % plot classifier

See also

datasets, mappings, scalem, klms, proxm, labeld, kernelc, logmlc,