Package de.jstacs.classifier.scoringFunctionBased.gendismix

Provides an implementation of a classifier that allows to train the parameters of a set of NormalizableScoringFunctions by a unified generative-discriminative learning principle

See:
          Description

Class Summary
GenDisMixClassifier This class implements a classifier the optimizes the following function
\[f(\underline{\lambda}|C,D,\underline{\alpha},\underline{\beta})
The weights $\beta_i$ have to sum to 1.
GenDisMixClassifierParameterSet This class contains the parameters for the GenDisMixClassifier.
LogGenDisMixFunction This class implements the the following function
\[f(\underline{\lambda}|C,D,\underline{\alpha},\underline{\beta})
The weights $\beta_i$ have to sum to 1.
OneSampleLogGenDisMixFunction This class implements the the following function
\[f(\underline{\lambda}|C,D,\underline{w},\underline{\alpha},\underline{\beta})
where $w_{c,n}$ is the weight for sequence $d_n$ and class $c$.
 

Enum Summary
LearningPrinciple This enum can be used to obtain the weights for well-known optimization tasks.
 

Package de.jstacs.classifier.scoringFunctionBased.gendismix Description

Provides an implementation of a classifier that allows to train the parameters of a set of NormalizableScoringFunctions by a unified generative-discriminative learning principle.

See Also:
NormalizableScoringFunction