get_sample_class_gmm (Operator)

Name

get_sample_class_gmm — Return a training sample from the training data of a Gaussian Mixture Models (GMM).

Synopsis

get_sample_class_gmm( : : GMMHandle, NumSample : Features, ClassID)

Description

get_sample_class_gmm reads out a training sample from the Gaussian Mixture Model (GMM) given by GMMHandle that was stored with add_sample_class_gmm or add_samples_image_class_gmm. The index of the sample is specified with NumSample. The index is counted from 0, i.e., NumSample must be a number between 0 and NumSamples - 1, where NumSamples can be determined with get_sample_num_class_gmm. The training sample is returned in Features and ClassID. Features is a feature vector of length NumDim, while ClassID is its class (see add_sample_class_gmm and create_class_gmm).

get_sample_class_gmm can, for example, be used to reclassify the training data with classify_class_gmm in order to determine which training samples, if any, are classified incorrectly.

Parameters

GMMHandle (input_control)	class_gmm → (integer)
GMM handle.

NumSample (input_control)	integer-array → (integer)
Index of the stored training sample.

Features (output_control)	real-array → (real)
Feature vector of the training sample.

ClassID (output_control)	number → (integer)
Class of the training sample.

Example

create_class_gmm (2, 2, [1,10], 'spherical', 'none', 2, 42, GMMHandle)
read_samples_class_gmm (GMMHandle, 'samples.gsf')
train_class_gmm (GMMHandle, 100, 1e-4, 'training', 1e-4, Centers, Iter)
* Reclassify the training samples
get_sample_num_class_gmm (GMMHandle, NumSamples)
for I := 0 to NumSamples-1 by 1
  get_sample_class_gmm (GMMHandle, I, Features, Class)
  classify_class_gmm (GMMHandle, Features, 2, ClassID, ClassProb,\
                      Density, KSigmaProb)
  if (not (Class=ClassProb[0]))
    * classified incorrectly
  endif
endfor
clear_class_gmm (GMMHandle)

Result

If the parameters are valid, the operator get_sample_class_gmm returns the value 2 (H_MSG_TRUE). If necessary an exception is raised.

Parallelization Information

get_sample_class_gmm is reentrant and processed without parallelization.

Possible Predecessors

add_sample_class_gmm, add_samples_image_class_gmm, read_samples_class_gmm, get_sample_num_class_gmm

Possible Successors

classify_class_gmm, evaluate_class_gmm

See also

create_class_gmm

Module

Foundation