CRF Class Reference

implements functions for training Conditional Random Fields More...

#include <merkmal.hh>

Static Public Member Functions
static void	train (Parameters *startPars, vector< AnnoSequence * > trainGenes, FeatureCollection &extrinsicFeatures)
static void	onlineLargeMarginTraining (Parameters *startPars, vector< AnnoSequence * > trainGenes, FeatureCollection &extrinsicFeatures)
static void	improvedIterativeScaling (Parameters *startPars, vector< AnnoSequence * > trainGenes, FeatureCollection &extrinsicFeatures)
template<class U , class V >
static void	compareWeights (Parameters *parameters, vector< U > &startWeights, vector< V > &endWeights, int numPrint=100)
static double	lossFct (Transcript *correct, Transcript *predicted)
static void	setEvalAnnoSeqs (vector< AnnoSequence * > seqs)
static void	setPrintPars (int ostatecount, StateModel **ostates)

Detailed Description

implements functions for training Conditional Random Fields

Author

Mario Stanke

---

The documentation for this class was generated from the following files:

• include/merkmal.hh
• src/merkmal.cc