WVOTEC

Weighted combiner (Adaboost weights)

   W = WVOTEC(A,V) compute weigths and store
   W = WVOTEC(V,U) Construct weighted combiner using weights U

   INPUT
     A Labeled dataset
     V Parallel or stacked set of trained classifiers
     U Set of classifier weights

   OUTPUT
     W Combined classifier

Description

The set of trained classifiers V is combined using weighted  majority voting. If given the weights U are used. If not  given, the weights are computed from the classification  results of the labeled dataset A using 0.5*log((1-E)/E)  if E is the classifier error.

See also

mappings, datasets,