Digital Signal Processing Reference
In-Depth Information
TWODPV=E2'*Ktest;
DETECTEDAS2D=[];
for i=1:1:12
[P,Q]= min(sum(repmat(TWODPV(:,i),1,4)-MEANVECTOR2').ˆ2);
DETECTEDAS2D=[DETECTEDAS2D Q];
end
POS2D=length(find((REF-DETECTEDAS2D)==0))/12;
In general the percentage of classification depends on the following.
1. Type of the feature vector.
2. Type of the projection technique.
3. Dimension of the projected vectors.
4. Number of training data.
5. Type of the classifier.
Search WWH ::
Custom Search