#include <Classifier.h>
Inheritance diagram for Impala::Core::Training::Classifier:
Public Member Functions | |
virtual | ~Classifier () |
virtual void | Train (Util::PropertySet *properties, TrainDataSrc *data)=0 |
virtual Table::ScoreTable * | Predict (TrainDataSrc *data)=0 |
virtual void | LoadModel (const std::string &name, Util::Database *db)=0 |
virtual void | SaveModel (const std::string &name, Util::Database *db)=0 |
Classifiers can Train() a model on TrainDataSrc with some properties. One can think of TrainDataSrc as a combination of annotation and feature data. Classifiers can also apply the model to data in Predict().
Currently it is possible to retrain a classifier many times. It might be more clear if classifier is the result of training, in that case training should be done in the constructor.
Definition at line 27 of file Classifier.h.