PRTools contents

PRTools manual

votec

VOTEC

Voting combining classifier

   W = VOTEC(V)
   W = V*VOTEC

Input
 V Set of classifiers

Output
 W Voting combiner

Description

If V = [V1,V2,V3,...] is a stacked set of classifiers trained for the  same classes, W is the voting combiner: it selects the class with the

highest vote of the base classifiers. This might also be used as
 A*[V1,V2,V3]*VOTEC in which A is a dataset to be classified.

The direct classifier outputs D = B*W for a test set B are posterior  probability estimates. D(i,j) = (v+1) / (n+c), in which v is the number  of votes object i receives for the j-th class. n is the total number of  classifiers, c the total number of classes.

The base classifiers may be combined in a stacked way (operating in the  same feature space) by V = [V1,V2,V3, ... ] or in a parallel way  (operating in different feature spaces) by V = [V1;V2;V3; ... ]

Example(s)

 PREX_COMBINING

See also

mappings, datasets, prodc, maxc, minc, medianc, meanc, averagec, stacked, parallel,

PRTools contents

PRTools manual