Updates the weights of the examples.
The re-weighting of 
(the weight vector over all the examples and classes) is done using the following formula
![\[ w^{(t+1)}_{i, \ell}= \frac{ w^{(t)}_{i, \ell} \exp \left( -\alpha^{(t)} h_\ell^{(t)}(x_i) y_{i, \ell} \right) }{ Z^{(t)} } \]](form_1.png)
where Z is a normalization factor, and it is defined as
![\[ Z^{(t)} = \sum_{j=1}^n \sum_{\ell=1}^k w^{(t)}_{j, \ell} \exp \left( -\alpha^{(t)} h_\ell^{(t)}(x_j) y_{j, \ell} \right) \]](form_2.png)
where 
is the number of examples, 
the number of classes, 
is the confidence in the weak classifier, 
is the classification of example 
for class 
with the classifier found at the current iteration (see BaseLearner::classify()), and 
is the binary label of that example, defined in InputData::getBinaryClass().
- Parameters:
-
pTrainData The pointer to the training data. pWeakHypothesis The current weak hypothesis.
- Returns:
- The value of the edge. It will be used to see if the algorithm can continue with learning.
- Date:
- 16/11/2005
Definition at line 207 of file AdaBoostMHLearner.cpp.
References BaseLearner::classify(), BaseLearner::getAlpha(), InputData::getBinaryClass(), ClassMappings::getNumClasses(), InputData::getNumExamples(), InputData::getWeight(), and InputData::setWeight().
