![]() |
Belle II Software
release-05-02-19
|
Generic expert for the regression applications. More...
#include <Regression.h>
Public Member Functions | |
void | load (Weightfile &weightfile) override |
Load the expert from the weightfile by loading each stored single classifier one ofter the other. | |
std::vector< float > | apply (Dataset &test_data) const override |
Apply the loaded experts by averaging over the single expert decisions. | |
virtual std::vector< float > | apply (Dataset &test_data, const unsigned int classID) const |
Apply this expert onto a dataset. More... | |
Protected Attributes | |
GeneralOptions | m_general_options |
General options loaded from the weightfile. | |
Private Attributes | |
std::vector< BaseClassifierExpert > | m_baseClassifierExperts |
The list of single experts. | |
Generic expert for the regression applications.
For each trained base classifier (for each bin, see the teacher for how this is done), one weight file is written as part into the full weight file. This expert reads all of them, instantiates a base expert for each bin and applies all of them to the dataset.
The result of the calculation is given by
probability = sum(expert probability for each expert) / # of experts
BaseClassifierExpert | Another expert that is related to the teacher you used for training |
RegressionClassifierOptions | The options of the classifier (probably inheriting from RegressionOptions) |
Definition at line 189 of file Regression.h.
|
inlinevirtualinherited |
Apply this expert onto a dataset.
Multi-class mode signature. Not pure virtual, since not all derived classes need to re-implement this.
test_data | dataset |
classID | class identifier. |
Reimplemented in TMVAExpertMulticlass.