FeatureTransformEstimateOptions Struct Reference

#include <get-feature-transform.h>

Collaboration diagram for FeatureTransformEstimateOptions:

Public Member Functions

 FeatureTransformEstimateOptions ()
 
void Register (OptionsItf *opts)
 

Public Attributes

bool remove_offset
 
int32 dim
 
BaseFloat within_class_factor
 
BaseFloat max_singular_value
 

Detailed Description

Definition at line 39 of file get-feature-transform.h.

Constructor & Destructor Documentation

◆ FeatureTransformEstimateOptions()

Member Function Documentation

◆ Register()

void Register ( OptionsItf opts)
inline

Definition at line 47 of file get-feature-transform.h.

References OptionsItf::Register().

Referenced by main().

47  {
48  opts->Register("remove-offset", &remove_offset, "If true, output an affine "
49  "transform that makes the projected data mean equal to zero.");
50  opts->Register("dim", &dim, "Dimension to project to with LDA");
51  opts->Register("within-class-factor", &within_class_factor, "If 1.0, do "
52  "conventional LDA where the within-class variance will be "
53  "unit in the projected space. May be set to less than 1.0, "
54  "which scales the features to have less variance, particularly "
55  "for dimensions where between-class variance is small. ");
56  opts->Register("max-singular-value", &max_singular_value, "If >0, maximum "
57  "allowed singular value of final transform (they are floored "
58  "to this)");
59  }

Member Data Documentation

◆ dim

◆ max_singular_value

BaseFloat max_singular_value

Definition at line 43 of file get-feature-transform.h.

Referenced by FeatureTransformEstimate::EstimateInternal().

◆ remove_offset

◆ within_class_factor

BaseFloat within_class_factor

Definition at line 42 of file get-feature-transform.h.

Referenced by FeatureTransformEstimate::EstimateInternal().


The documentation for this struct was generated from the following file: